Developmental Biology Protocols: Volume II [Volume 2, 1 ed.] 9780896038530, 9781592590650, 089603853X

Citation preview

Methods in Molecular Biology

TM TM

VOLUME 136

Developmental Biology Protocols Volume II Edited by

Rocky S. Tuan Cecilia W. Lo

HUMANA PRESS

Developmental Biology Protocols: Overview II

3

1 Developmental Biology Protocols Overview II Rocky S. Tuan and Cecilia W. Lo 1. Introduction The discipline of developmental biology covers scientific investigations aimed at deciphering the underlying mechanisms responsible for diversity and order within tissues and organs. It is the goal of Developmental Biology Protocols to present to the readers a set of contemporary, practical experimental tools dealing with a wide-ranging spectrum of topics in developmental biology research. This second volume of the three-volume set begins by presenting the tissue and organ models currently being studied, and the characteristics of abnormal development. The volume concludes with detailed description of the technologies used to identify developmentally important genes and the methods for transgenesis, including gene knockout. 2. Organogenesis Studies on Drosophila (Chapter 2) and Xenopus (Chapter 3) are first presented to illustrate the power of these systems for deciphering the experimental principles of developmental biology. Practical details on a number of organ/tissue systems in vertebrates are presented, including mammary gland (Chapter 4), heart (Chapter 5), skeleton (craniofacial: Chapters 6 and 7; axial and appendicular, Chapter 8), limb (Chapter 9), thymus (Chapter 10), liver (Chapter 11), and skin (Chapter 12). These systems exemplify the diversity of developmental mechanisms involved in organogenesis, as well as the experimental techniques applicable to their analysis. In addition, because programmed cell death has emerged as a common mechanistic step in many aspects of morphogenesis, two chapters are devoted to the methods of the analysis of apoptosis (Chapters 13 and 14). 3. Abnormal Development and Teratology Birth defects are the leading cause of infant mortality and are responsible for substantial morbidity and disability. In the United States alone, over 120,000 babies are born each year with a structural birth defect or malformation. At present, the causes remain unknown for most of these cases. One of the long-term goals of the science of developmental biology is, indeed, the discovery of the mechanisms responsible for From: Methods in Molecular Biology, Vol. 136: Developmental Biology Protocols, Vol. II Edited by: R. S. Tuan and C. W. Lo © Humana Press Inc., Totowa, NJ

3

4

Tuan and Lo

abnormal development. This section contains chapters representing a cross-section of many such examples and the experimental approaches currently used in analyzing the underlying mechanisms. Experimental studies of neurulation and neural tube defects are first described (Chapters 16–19). Abnormalities in placentation, which is critical for a range of physiological interactions required for fetal growth and development, are often associated with early embryonic mortality as well as serious pregnancy disorders such as pre-eclampsia; methods for examining placentation is presented in Chapter 20. Craniofacial malformations, such as cleft palate, are among the most frequent birth defects in live-born human infants, and two chapters (Chapters 21 and 22) are devoted to the analysis of palatal dysmorphogenesis. The method of interspecies tissue grafting, also detailed in Volume I of this series, is described here in the context of understanding the basis of developmental limb anomalies (Chapter 23). Techniques used to study cardiac morphogenesis and dysmorphogenesis, particularly related to laterality defects, are detailed in Chapters 24–26. Finally, in assessing the developmental toxicity of potentially harmful substances, it is crucial to utilize valid dose-response models; Chapter 27 summarizes the biologically based risk assessment models for developmental toxicity. 4. Screening and Mapping of Novel Genes and Mutations The utility of applying molecular biology techniques to the study of development is well documented in this Part. Specifically, developmental biologists have successfully adopted differential gene screening and cloning approaches to relate specific gene expression events to spatiotemporally defined stages of development and embryogenesis, such as cellular commitment, differentiation, and morphogenesis. Such correlations then provide a rational basis for further analysis of the putative functions of specific genes and their products. Examples of differential screening of developmental gene expression are presented in Chapters 31–33. Gene cloning may be carried out by positional cloning (Chapter 28) and gene trapping in embryonic stem cells (Chapter 29). In most cases, cloning is accomplished using polymerase chain reaction (PCR)-based methods, as described in Chapter 30. 5. Transgenesis: Production and Gene Knockout The development of techniques that permit the introduction of gene sequences into the zygote or embryonic stem cells to produce transgenic animals ranks as one of the seminal achievements of the molecular biology revolution. The chapters in this Part provide a panoramic portfolio of the many diverse and powerful approaches that have been developed to produce transgenic animals and the manipulations that permit targeted gene knockouts. Protocols used in producing transgenic Drosophila (Chapter 34), sea urchin (Chapter 35), and zebrafish (Chapter 36) are first described. Recent technical advances in the production of avian transgenics are also summarized (Chapters 37–39). Methods for the production of mammalian, particularly mouse, transgenic animals are described in Chapters 40–42, the last dealing specifically with the use of yeast artificial chromosomes (YACs). A novel method for high-efficiency formation of chimeric animals with germline gene transmission is presented in Chapter 44. Finally, exciting recent developments in gene-targeting strategies (Chapter 43), and the application of the Cre/LoxP site-specific recombination for conditional gene knockout (Chapter 45) and transgene coplacement for Drosophila (Chapter 46) are also presented.

Drosophila as a Genetic Tool

7

2 Drosophila as a Genetic Tool to Define Vertebrate Pathway Players Nancy M. Bonini

I. Introduction In many instances, the strength of Drosophila melanogaster genetics can be used to enhance our understanding of complex vertebrate signaling systems. The general success of this approach is underscored by the large number of vertebrate signaling components whose very names derive in part from the names of Drosophila mutants. Examples include the vertebrate pathway components Sonic Hedgehog, Son of Sevenless, Lunatic Fringe, Notch, the SMAD family of transforming growth factor-β (TGF-β) signaling, and many others. Given the powerful genetics of Drosophila melanogaster (see ref. 1), it can be of interest to test functional equivalence of vertebrate homologs with fly genetic pathway components, or to re-create in Drosophila transgenic models for vertebrate or human gene function. If such complementation can be established, then the strength of Drosophila genetics can be brought to bear on defining additional components of the particular pathway of interest; for example, through enhancer and suppressor screens. Subsequently, one can then clone such modifier genes from Drosophila, as a springboard from which to identify their vertebrate counterparts. To establish a genetic model for a vertebrate gene function in Drosophila, there are a number of considerations with respect to expressing foreign genes in the fly, establishing whether and how the foreign proteins function, and using the transgenic lines in genetic screens. Examples of functional complementation in flies with vertebrate genes include the ability of domains of human bone morphogenetic protein to substitute in the related fly protein Dpp (2), effects of vertebrate fringe homologs to establish boundaries like the fly gene (3), functional complementation of orthodenticle homeobox gene homologs (4–6), and functional complementation of mammalian counterparts of eye determination genes eyeless and eyes absent (7,8). In addition, dominant effects can be generated in flies with vertebrate genes, such as phenocopying fly homeotic mutants with the appropriate vertebrate Hox homologs (9,10), and generating genetically tractable human disease models by expressing mutant human disease proteins in flies (11).

From: Methods in Molecular Biology, Vol. 136: Developmental Biology Protocols, Vol. II Edited by: R. S. Tuan and C. W. Lo © Humana Press Inc., Totowa, NJ

7

8

Bonini

2. Experimental Approaches

2.1. Design of Constructs for Transformation As noted, there are a number of examples of expressing vertebrate cDNA counterparts of a particular gene of interest in flies. In general, it appears that simply taking a human, mouse, or other vertebrate cDNA and expressing it in flies will usually generate a functional protein. Thus, it has not proven necessary to be overly concerned about possible differential codon usage from flies to vertebrates. However, because the insect body plan is quite different from the typical mammalian body plan, gene regulatory sequences cannot be transferred between species so easily (but see refs. 10 and 12). Consequently, care must be taking to select an expression system compatible with Drosophila to achieve adequate protein levels in the relevant tissues, as discussed below. An important consideration is the ability to detect the foreign protein when expressed in flies, especially when one considers that the Drosophila genome displays position effects such that some transgenic insertions will express at higher levels than others as a result of the location in which the transgene has inserted in the genome (13). Transformation vectors can include insulators, such that the transgene will be much less sensitive to genomic position effects (14). Otherwise, it is typically necessary to generate a number of different transgenic lines in order to obtain a sufficient number with strong expression; a minimum estimate is about four lines. Having different lines that express at weak, moderate, or strong levels to give a weak, moderate, or strong phenotype, respectively, can be of benefit, however, especially when performing genetic screens for modification of the phenotype (e.g., see refs. 15–17). If the experiments require the construction of a large number of genetic stocks, having the ability to detect expression of the foreign protein can be extremely valuable to allow selection of strongly expressing transgenic lines. In addition, if a phenotype is not observed, unless transgene expression can be monitored directly, it might be difficult to distinguish whether the foreign protein does not function in flies or if there is simply a technical problem with expression. It is of course possible to use in situ hybridization to detect expression of the transcript for the transgene. A disadvantage of this approach is that it is, in general, more laborious than detecting protein expression and, moreover, does not indicate whether the protein is being translated appropriately in the fly. In some cases, an antibody to the foreign protein may already be available; one can then test for crossreactivity to potential fly counterparts to determine the utility and limitations of the antibody. When testing for antibody crossreactivity, it is frequently necessary to preadsorb an antibody against fixed fly tissue (e.g., a 1:10 antibody dilution preadsorbed with 50 µL of 4% paraformaldehyde-fixed, dechorionated, devitellinized embryos) to lower potential background crossreactivity. This is particularly necessary for rabbit antisera, which are notorious for giving a high background on fly tissue. It is also necessary to determine whether the antibody to the vertebrate counterpart crossreacts to the fly counterpart; if so, one must be able to distinguish expression of the vertebrate counterpart from the fly gene by some other means, such as expression in a novel tissue where the fly gene is not normally expressed or by tagging the vertebrate protein with a peptide domain to which antibodies are available. An alternative approach is to tag the foreign protein with a small peptide sequence for which antibodies are available. Examples include FLAG, c-Myc, and hemaglutinin

Drosophila as a Genetic Tool

9

(HA), for which antibodies can be purchased commercially; alternatively, fusion to a protein with endogenous fluorescence, such as green fluorescent protein (GFP) or one of its derivatives, can be used. If electron microscopy is ultimately of interest, then a glutaraldehyde-resistant epitope tag is particularly useful, as many antibodies lose reactivity to tissue treated with glutaraldehyde [although we have successfully performed immunoelectron microscopy with HA-tagged protein (11)]. Such epitope tags can be added by polymerase chain reaction (PCR) or by subcloning into various commercially available vectors which have these epitopes upstream of a number of convenient restriction enzyme sites. We have added HA and GFP to the N- or C-terminus of a number of proteins successfully (18). Frequently, we add HA to the C-terminus using PCR. To do this, we design a C-terminal primer which deletes the stop codon and adds a linker of a few small amino acids (glycine and alanine, plus a convenient restriction site), followed by the HA sequence and a stop codon. One can also multimerize the exogenous tag (3–5×) to boost sensitivity of detection (19). If the tag is at the C-terminus, then one can be assured that the entire protein is being produced if the introduced protein can be detected with the relevant antibody. Alternatively, Western immunoblotting can be used to confirm the synthesis of a protein of the appropriate size.

2.2. Expression Systems There are a number of different expression systems available, the simplest of which couples a standard transformation vector with an appropriate promoter. Such a promoter may be conditional, such as a heat-shock promoter which is inducible by heat pulsing the animal at 37°C for a short time (20). Alternatively, it may be a constitutive promoter expressed in a tissue of interest, such as actin or ubiquitin which will be expressed in most cells of the animal (21), or a promoter that targets gene expression to a particular tissue, such as the gmr (glass multiple reporter) or sevenless promoter elements which target gene expression to developing eye cells (22–24). Such constructs have the advantages of simplicity and, depending on the promoter used, yield a transgenic line with a constant and consistent phenotype. Conditional promoters allow one to express the protein at any desired time; however, in general, expression will vary over time (although the heat-shock promoter can give a constant basal level of expression at normal growth temperatures, depending on insertion site, which can be sufficient for a phenotype at normal growth temperatures [e.g., ref. 25]). If one suspects that ubiquitous or early expression of the protein may be lethal to the animal, then conditional or tissue-restricted expression is essential. Another approach is a two-component system, the GAL4-UAS system (26). In this system, the gene of interest is cloned downstream of the yeast UAS–GAL4 DNA-binding regulatory sequences in a fly transformation vector pUAST, and transgenic lines are generated. Then, upon crossing the transgenic line to any of a large collection of fly lines that express GAL4 in tissue-specific patterns, one can express the gene of interest in different tissues at different times of development. One advantage of this system is versatility, as there are many GAL4 lines with different expression patterns available from Drosophila stock centers or research laboratories. In this system, a UAS–lacZ tester strain can be used to monitor promoter strength and tissue-specific expression of the GAL4 lines being used. In addition, an advantage is the ease of determining the viability or other features of the phenotype—even if expressing the protein widely is

10

Bonini

lethal, one may be able to obtain transgenic lines because expression is only induced when the transgenic line is crossed to a GAL4 expression line. Conversely, the fact that crosses must be made in order to express the transgene represents a disadvantage of the GAL4–UAS system. Furthermore, the double-insert line of interest is not of itself stable unless one takes the trouble to generate an appropriate stable recombinant fly line. This requirement can become particularly unwieldy when testing the phenotype of a foreign gene in a fly mutant background—performing a single experiment can require many crosses to assemble a complex combination of mutant alleles and transgenic constructs. Again, one must consider the different potential uses of the transgenic line in the long run to determine which approach or approaches will be best suited for the experiments.

2.3. Testing for Function There are a number of ways to test for function of a foreign protein in transgenic Drosophila. If testing homologs of a known fly gene for which mutants exist, then one test for function is ability of the foreign gene to rescue the fly mutant phenotype. If the fly counterpart has dominant effects or if one might expect dominant effects as a result of the function of the protein in vertebrates (such as for a dominant oncogene or disease gene), then another test is to determine whether the vertebrate homolog can induce similar dominant phenotypes in flies. There are examples of dominant oncogenic mutations leading to a form of the protein that also functions dominantly in the fly (27–33). In some cases, expression of vertebrate genes in flies has demonstrated that a conserved function of the vertebrate and fly genes is autoregulation; thus, the vertebrate protein (frequently a transcription factor) turns on expression of the endogenous fly counterpart (9,34). If one has mutants in the fly gene involved, then it is possible to test for functional conservation in the genetic background of a protein null of the fly gene and, hence, address broader aspects of functional conservation (e.g., ref. 5). When expressing a foreign gene in the fly in a tissue that normally does not express any such gene, one must consider if screens to identify interacting proteins will be useful for understanding the function of the gene in its normal cellular context. It is important to assess whether any phenotypic effects observed in the fly accurately reflect conserved functions of the vertebrate protein under scrutiny. For example, will vertebrate anti-apoptotic genes block Drosophila programmed cell death? Will the vertebrate homolog, like its fly counterpart, direct ectopic tissue formation in the fly? If the vertebrate cDNA induces a dominant effect, is that effect the result of elevated levels of a normal activity of the protein (a hypermorphic effect) or of a new activity of the protein that may have little to do with its normal function (a neomorphic effect). Neomorphic effects, for example, might be the result of subcellular mislocalization of the vertebrate protein in the fly. To what degree does the pathology of a human disease gene reflect biological effects known to occur in humans or vertebrate models, and can these effects be faithfully replicated in the fly model? These are, of course, specific issues that vary for any one gene of interest, and they are critical to consider.

2.4. Genetic Screens for Modifier Mutations A major goal of expressing a foreign protein in flies is to be able to apply Drosophila genetics to further understand the biological problem. The basic idea is to find mutations in fly genes that enhance or suppress the phenotype, and use these mutations

Drosophila as a Genetic Tool

11

to identify vertebrate genes that function in the same pathways or biological process. By this means, one can, therefore, define additional genes that elucidate or indirectly influence the biological pathway of interest. There are two general approaches for identifying modifier mutations: (1) to screen collections of existing mutations or deficiencies to define interacting genes and (2) do a de novo mutagenesis in flies to define interacting genes. Usually, both approaches are performed as screens for dominantly modifying mutations on the autosomes and recessive or dominant mutations on the X chromosome. These screens allow direct analysis of modifying effects in the progeny of mass fly matings, enabling a large number of potential mutants to be rapidly screened relative to other methods. One approach is to look for enhancers or suppressors by crossing the flies bearing the foreign gene of interest to a collection of Drosophila deficiency chromosomes. This collection, available from the Bloomington Drosophila Stock Center, consists of about 190 fly lines, which uncover, in total, approximately 70–80% of the Drosophila euchromatin. By this approach, one searches for regions of the chromosomes that harbor genes that, when reduced in dosage by 50%, will modify the phenotype of interest. Thus, to test all regions of the genome uncovered by available deficiencies, one simply performs fly crosses and examines the resulting progeny flies. Once a deficiency region of interest is found, then the genetic interaction can be confirmed and the cytological region of the chromosome narrowed down as much as possible using smaller available deficiencies. Eventually, one can test for interactions with all available known mutations in the region and/or perform a mutagenesis to define genes in the region. Hay et al. (15) have successfully used this strategy to identify a conserved gene that is involved in programmed cell death pathways. A disadvantage of this technique is that the deficiency lines tend to show variable genetic background effects; that is, it is difficult to determine whether any observed effect on the phenotype of interest is the result of the deficiency itself or to the fact that the cross is made between nonisogenic fly lines. Thus, the success of the approach can depend on the strength and variability of the phenotype being modified. If the modifier effect is very strong, then this approach can be quite successful; however, if the modifier effect is subtle, then it can be difficult to distinguish modification of the phenotype in the widely variable backgrounds of the deficiency lines. Another disadvantage is that eventually after narrowing down a region to the smallest possible extent, it may still be necessary to perform many molecular biological manipulations before having a defined gene in hand. A variation on this approach is to look for modifier mutations among the large collection of P-element-induced mutation lines (36,37). Should an interaction be found by using P-element-induced mutations to look for dosage-sensitive modifier interactions, then the gene can easily be cloned if the P element has inserted into it or nearby. In addition, some of the P lethals have been generated using reporter gene constructs, such that one can stain the line for the reporter gene expression (β-galactosidase), which may reveal interesting aspects of the expression pattern of the potentially interacting gene. However, the P-element-induced mutations have a similar background problem as the deficiency lines, which, again, can often be too variable in practice to make such a screen successful (e.g., see ref. 17). In general, both deficiencies and the P lethals test for the same type of interaction: an interaction resulting from reduction of a gene dosage by 50%. An alternative approach

12

Bonini

is to use a point mutagen (ethyl methanesulfonate [EMS] is commonly used in Drosophila) to identify dominantly interacting mutations. With a point mutagen, one can obtain both loss-of-function mutations that reduce gene function and gain-of-function mutations resulting from single amino acid substitution in a critical region of the interacting protein. Thus, using EMS as a mutagen may select for different types of interacting mutations than a deficiency or P-element lethal screen. X-ray mutagenesis can also be of interest, because X-rays will, in general, produce chromosomal rearrangements that can affect very large genes or gene complexes as well as result in gain-offunction mutations, depending on the particular rearrangement (see ref. 17). Most crucially, by doing such a mutagenesis, one has greater control over the genetic background: one can select an isogenic background that, when crossed to the line of interest, gives a uniform phenotype such that the effect of any modifier interaction will be readily seen. Such an approach has proven successful for a number of different types of modifier screens (16,17). By any of these approaches, the real challenge comes in the analysis of the modifiers obtained to identify those that are most interesting with respect to the question of interest. In all of these approaches, it is essential to have good controls to eliminate modifiers that interact with the expression system or the promoter expressing the gene rather than with the protein being expressed, and so forth. Thus, secondary screens are critical to classify mutants to distinguish those modifiers more directly involved in the question of interest, from those that are only peripherally involved. An excellent example of this is ref. 38, where 30,000 mutagenized chromosomes were screened for modification of a sevenless receptor tyrosine kinase mutant phenotype. Of seven complementation groups identified, four of seven also modified the mutant phenotype of a second tyrosine kinase receptor (the EGF receptor), thus defining those genes that were common signaling components of receptor tyrosine kinase pathways. Some argue that the best approach is to do different types of modifier screens and then focus on those subsets of new genes that are repeatedly identified in multiple screens, indicating that they are likely to be centrally important in the biological pathway of interest. By these means, one can apply the ease and rapidity of genetics in a simple model system like Drosophila to questions of fundamental interest and importance in vertebrates. With the advent of genomic sequencing, the importance of model systems like Drosophila to reveal protein function and define biological pathways becomes ever more important. Acknowledgments Thanks to A. Cashmore and M. Fortini for critical comments. N.B. is supported by grants from the John Merck Fund, the National Eye Institute, the Alzheimer’s Association, and the David and Lucile Packard Foundation. References 1. Rubin, G. M. (1988) Drosophila melanogaster as an experimental organism. Science 240, 1453–1459. 2. Padgett, R. W., Wozney, J. M., and Gelbart, W. M. (1993) Human BMP sequences can confer normal dorsal-ventral patterning in the Drosophila embryo. Proc. Natl. Acad. Sci. USA 90, 2905–2909.

Drosophila as a Genetic Tool

13

3. Johnston, S. H., Rauskolb, C., Wilson, R., Prabhakaran, B., Irvine, K. D., and Vogt, T. F. (1997) A family of mammalian Fringe genes implicated in boundary determination and the Notch pathway. Development 124, 2245–2254. 4. Acampora, D., Avantaggiato, V., Tuorto, F., Barone, P., Reichert, H., Finkelstein, R., and Simeone, A. (1998) Murine Otx1 and Drosophila otd genes share conserved genetic functions required in invertebrate and vertebrate brain development. Development 125, 1691–1702. 5. Leuzinger, S., Hirth, F., Gerlich, D., Acampora, D., Simeone, A., Gehring, W. J., Finkelstein, R., Furukubo-Tokunaga, K., and Reichert, H. (1998) Equivalence of the fly orthodenticle gene and the human OTX genes in embryonic brain development of Drosophila. Development 125, 1703–1710. 6. Nagao, T., Leuzinger, S., Acampora, D., Simeone, A., Finkelstein, R., Reichert, H., and Furukubo-Tokunaga, K. (1998) Developmental rescue of Drosophila cephalic defects by the human Otx genes. Proc. Natl. Acad. Sci. USA 95, 3737–3742. 7. Halder, G., Callaerts, P., and Gehring, W. (1995) Induction of ectopic eyes by targeted expression of the eyeless gene of Drosophila. Science 267, 1788–1792. 8. Bonini, N. M., Bui, Q. T., Gray-Board, G. L., and Warrick, J. M. (1997) The Drosophila eyes absent gene directs ectopic eye formation in a pathway conserved between flies and vertebrates. Development 124, 4819–4826. 9. McGinnis, N., Kuziora, M. A., and McGinnis, W. (1990) Human Hox-4.2 and Drosophila deformed encode similar regulatory specificities in Drosophila embryos and larvae. Cell 63, 969–976. 10. Malicki, J., Cianetti, L. C., Peschle, C., and McGinnis, W. (1992) A human HOX4B regulatory element provides head-specific expression in Drosophila embryos. Nature 358, 345–347. 11. Warrick, J. M., Paulson, H., Gray-Board, G. L., Bui, Q. T., Fischbeck, K., Pittman, R. N., and Bonini, N. M. (1998) Expanded polyglutamine protein forms nuclear inclusions and causes neural degeneration in Drosophila. Cell 93, 939–949. 12. Haerry, T. E. and Gehring, W. J. (1996) Intron of mouse Hoxa-7 gene contains conserved homeodomain binding sites that can function as an enhancer in Drosophila. Proc. Natl. Acad. Sci. USA 93, 13,884–13,889. 13. Spradling, A. C. and Rubin, G. M. (1983) The effect of chromosomal position on the expression of the Drosophila xanthine dehydrogenase gene. Cell 34, 47–57. 14. Kellum, R. and Schedl, P. (1991) A position-effect assay for boundaries of higher order chromosomal domains. Cell 64, 941–950. 15. Hay, B. A., Wassarman, D. A., and Rubin, G. M. (1995) Drosophila homologs of baculovirus inhibitor of apoptosis proteins function to block cell death. Cell 83, 1253–1262. 16. Dickson, B. J., van der Straten, A., Dominguez, M., and Hafen, E. (1996) Mutations modulating Raf signaling in Drosophila eye development. Genetics 142, 163–171. 17. Karim, F. D., Chang, H. C., Therrien, M., Wassarman, D. A., Laverty, T., and Rubin, G. M. (1996) A screen for genes that function downstream of Ras1 during Drosophila eye development. Genetics 143, 315–329. 18. Bui, Q. and Bonini, N., unpublished results. 19. Panin, V., Papayannopoulos, V., Wilson, R., and Irvine, K. (1997) Fringe modulates Notch-ligand interactions. Nature 387, 908–912. 20. Schneuwly, S., Klemenz, R., and Gehring, W. (1987) Redesigning the body plan of Drosophila by ectopic expression of the homeotic gene Antennapedia. Nature 325, 816–818. 21. Sekelsky, J. J., Newfeld, S. J., Raftery, L. A., Chartoff, E. H., and Gelbart, W. M. (1995) Genetic characterization and cloning of mothers against dpp, a gene required for decapentaplegic function in Drosophila melanogaster. Genetics 139, 1347–1358.

14

Bonini

22. Basler, K., Siegrist, P., and Hafen, E. (1989) The spatial and temporal expression pattern of sevenless is exclusively controlled by gene-internal elements. EMBO J. 8, 2381–2386. 23. Bowtell, D. D., Lila, T., Michael, W. M., Hackett, D., and Rubin, G. M. (1991) Analysis of the enhancer element that controls expression of sevenless in the developing Drosophila eye. Proc. Natl. Acad. Sci. USA 88, 6853–6857. 24. Ellis, M., O’Neill, E., and Rubin, G. (1993) Expression of Drosophila glass protein and evidence for negative regulation of its activity in non-neuronal cells by another DNAbinding protein. Development 119, 855–865. 25. Fehon, R. G., Oren, T., LaJeunesse, D. R., Melby, T. E., and McCartney, B. M. (1997) Isolation of mutations in the Drosophila homologues of the human Neurofibromatosis 2 and yeast CDC42 genes using a simple and efficient reverse-genetic method. Genetics 146, 245–252. 26. Brand, A. H. and Perrimon, N. (1993) Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development 118, 401–415. 27. Bishop, J. G., III and Corces, V. (1988) Expression of an activated ras gene causes developmental abnormalities in transgenic Drosophila melanogaster. Genes Dev. 2, 567–577. 28. Fortini, M., Simon, M., and Rubin, G. (1992) Signalling by the sevenless protein tyrosine kinase is mimicked by Ras1 activation. Nature 355, 559–561. 29. Fortini, M. E., Rebay, I., Caron, L. A., and Artavanis-Tsakonas, S. (1993) An activated Notch receptor blocks cell-fate commitment in the developing Drosophila eye. Nature 365, 555–557. 30. Lieber, T., Kidd, S., Alcamo, E., Corbin, V., and Young, M. W. (1993) Antineurogenic phenotypes induced by truncated Notch proteins indicate a role in signal transduction and may point to a novel function for Notch in nuclei. Genes Dev. 7, 1949–1965. 31. Rebay, I., Fehon, R. G., and Artavanis-Tsakonas, S. (1993) Specific truncations of Drosophila Notch define dominant activated and dominant negative forms of the receptor. Cell 74, 319–329. 32. Struhl, G., Fitzgerald, K., and Greenwald, I. (1993) Intrinsic activity of the Lin-12 and Notch intracellular domains in vivo. Cell 74, 331–345. 33. Casanova, J., Llimargas, M., Greenwood, S., and Struhl, G. (1994) An oncogenic form of human raf can specify terminal body pattern in Drosophila. Mech. Dev. 48, 59–64. 34. Glardon, S., Callaerts, P., Halder, G., and Gehring, W. (1997) Conservation of Pax-6 in a lower chordate, the ascidian Phallusia mammillata. Development 124, 817–825. 35. Hay, B. A., Wolff, T., and Rubin, G. M. (1994) Expression of baculovirus P35 prevents cell death in Drosophila. Development 120, 2121–2129. 36. Cooley, L., Kelley, R., and Spradling, A. (1988) Insertional mutagenesis of the Drosophila genome with single P elements. Science 239, 1121–1128. 37. Török, T., Tick, G., Alvarado, M., and Kiss, I. (1993) P-lacW insertional mutagenesis on the second chromosome of Drosophila melanogaster: isolation of lethals with different overgrowth phenotypes. Genetics 135, 71–80. 38. Simon, M. A., Bowtell, D. D. L., Dodson, G. S., Laverty, T. R., and Rubin, G. M. (1991) Ras1 and a putative guanine nucleotide exchange factor perform crucial steps in signaling by the Sevenless protein tyrosine kinase. Cell 67, 701–716.

Activin in Amphibian Embryogenesis

15

3 Bioassays for Studying the Role of the Peptide Growth Factor Activin in Early Amphibian Embryogenesis Makoto Asashima, Takashi Ariizumi, Shuji Takahashi, and George M. Malacinski 1. Introduction Activin, a peptide growth factor, is a member of the transforming growth factor-β (TGF-β) superfamily. It was originally isolated from follicle fluid as a gonadal hormone that stimulates follicle-stimulating hormone (FSH) secretion and is identical to EDF, the erythroid differentiation factor (which stimulates erythroleukemia cells to differentiate into hemoglobin-producing cells) (1). It also appears to be related, if not identical, to the so-called “vegetalizing factor” described originally by Tiedemann and colleagues, which can induce amphibian embryonic tissue rudiments to display various differentiation patterns (reviewed in ref. 2). In addition, it appears to be identical to the XTC factor isolated by Smith et al. (3) from transformed Xenopus fibroblasts. Thus, the identification of activin as a potential morphogen in amphibian embryos (4) solved several mysteries surrounding the puzzle regarding the molecular nature of various hitherto ill-characterized “inducing substances.” Activin is a dimeric protein and has been described in mammals as sharing subunits with a related dimeric protein—inhibin (Fig. 1). Inhibin is another peptide, which, in contrast to activin, inhibits the secretion of FSH by pituitary cells. Although inhibins have been identified in mammals, they have not yet been described in amphibia. Activin warrants nomination as a candidate endogenous inducer, for it is found in embryos at the expected place and at the appropriate time to play a role in mesoderm formation. For example, activin proteins have been detected in unfertilized eggs (5,6) and have been demonstrated to be bound to the vitellogenin component of yolk platelets (7). We know that the maternal supply of activin protein is most likely synthesized outside of the oocyte (e.g., in follicle cells) and transported into the oocyte, because activin mRNA has been localized to follicle cells (8,9). It may actually be transported “piggyback” fashion into the oocyte on yolk proteins which are synthesized in the liver during the vitellogenic stage of oogenesis. Later, during the blastula/gastrula stage, activin genes are expressed from the zygotic genome (10). From: Methods in Molecular Biology, Vol. 136: Developmental Biology Protocols, Vol. II Edited by: R. S. Tuan and C. W. Lo © Humana Press Inc., Totowa, NJ

15

16

Asashima et al.

Fig. 1. Molecular structure of activin and inhibin proteins. Activin A is a homodimer consisting of two inhibin βA chains; activin AB is a heterodimer of inhibin βA and βB chains; and activin B is a homodimer of two inhibin βB subunits. All three forms have been identified in amphibian embryos (see ref. 5). The native molecular weight of activin is 25,000 D.

The potential importance of activin studies is substantial: Activin induces mesoderm development, and mesoderm induction appears to be a common early feature of the patterning of virtually all vertebrate embryos. Thus, to be able to study the mechanism of action of a purified protein that can generate simulated mesoderm induction in vitro is a major experimental advantage. Indeed, a wide range of mesodermal tissues can be induced with activin and profound effects on embryonic patterning can be generated, depending upon the stage of the responding tissue, concentration of activin (11,12), and type of bioassay. For example, in the tissue-culture assay to be described below, with relatively low concentrations of activin (e.g., 0.3 ng/mL), only ventral mesoderm such as blood cells and coelomic epithelium is induced, whereas at 5–10 ng/mL, muscle is induced, and at higher concentrations (e.g., 50 ng/mL), notochord (a distinctly dorsal structure) is induced. A range of dramatic patterning effects can also be observed using one of the injection assays described below. Consideration of those bioassay results makes it easy to conceive of a model for pattern formation that uses varying concentrations of activin as a key molecule in concentration gradient-dependent patterning scenarios. As well, models for patterning that involve serial inductions initiated by activin action can be readily conceptualized. When activin-treated newt animal cap is sandwiched between nontreated animal caps, an induction cascade reminiscent of classical Spemann-embryonic organizer action is observed (13). The type of axial tissue induced (i.e., head vs tail) depends on the dura-

Activin in Amphibian Embryogenesis

17

tion of the preculture period of activin-treated animal cap. This observation is also reminiscent of the original primary embryonic organizer grafting experiments, which demonstrated that early-stage organizer tissue induces head structures, whereas a laterstage organizer induces mostly tail tissue. Verification of the nature of the induced tissues can be achieved with the use of either histological or molecular indices of gene expression (see below). In addition to using activin’s cell differentiation effects as a model system for understanding the molecular basis of patterning in the early embryo, activin can be employed either alone or in combination with other prospective regulatory molecules for two other purposes as well: (1) studying differentiation of specific tissues/organs in culture [e.g., renal tissue (14) and heart (15)] and (2) understanding theoretical aspects of embryonic development. Concerning this latter point, we are presently initiating an ambitious undertaking, the so-called A-FEAT project: an attempt to assemble a functioning embryo from activin-treated components. It is briefly described in Subheading 4. 2. Types of Bioassay Protocols Several different bioassay schemes for activin studies have been developed. Each measures a different aspect of the action of activin. The tissue-culture assays, for example, focus on the effects of activin on cell differentiation. The injection assays are designed to recognize activin effects on embryonic patterning. Finally, the interference–injection assays target endogenous activin mechanisms. Each will be described in turn.

2.1. Tissue-Culture Assay This assay protocol is perhaps the simplest, for it does not require expertise in microinjection techniques. However, as will be mentioned later, it is important that the preparation of the explanted tissue be carefully monitored. Essentially, animal cap tissue is bathed in a simple salt (culture) medium containing an appropriate concentration of activin. Nontreated blastomeres differentiate into “atypical epidermis,” whereas treated animal cap differentiates into ventral mesodermal cell types, such as blood, mesenchyme, and coelomic epithelium, or into dorsal mesoderm, such as muscle and notochord. Figure 2 illustrates the general experimental protocol for using activin to generate that range of cell differentiations in cultured animal cap tissue. Scoring the results can be accomplished by either observing histological indices of differentiation (Fig. 3) or by detecting the appearance of specific molecular markers (Table 1). Thus, this bioassay can be employed to determine the extent to which activin promotes the development of specific differentiation profiles in cultured tissues. Although it has been used primarily with Xenopus test systems, we have also used it often for analyzing organogenesis in urodele animal cap tissue. For example, urodele animal caps, when treated with high concentrations of activin, developed beating hearts (15). This tissue-culture assay offers several advantages to the experimenter. Primarily, of course, it is relatively easy to perform. The culturing of amphibian tissues is remarkably simple because early-stage embryonic tissue is preloaded with an abundant supply

18

18 Asashima et al.

Fig. 2. The animal cap assay generates a broad range of cell differentiations, depending on the concentration of activin in the culture medium. Typically, a Xenopus animal cap is cultured in the presence of activin for 3 d (20°C). Low concentrations (e.g., 0.5 ng/mL) induce ventral mesodermal tissue. Mid-range concentrations (e.g., 10 ng/mL) induce muscle, whereas higher concentrations (e.g., 50 ng/mL) induce dorsal mesodermal structures (e.g., notochord). IL11: interleukin 11; SCF: stem cell factor; RA: retinoic acid.

Activin in Amphibian Embryogenesis

19

Fig. 3. Histological examination of Xenopus animal caps. treated with activin for 3 d. Using the fixation and staining methods described in Subheading 3. differentiated tissues were easily recognized. These included blood cells (bl) and coelomic epithelium (ce) (A), muscle (mus) (B), and notochord (not) (C). A section through an animal cap that received no activin treatment (control) is shown in (D).

of nutrients, so the culture medium need only consist of a simple buffered salt solution. Purified activin is available and its concentration in a culture medium is easily adjusted. It is also easy to add additional factors to the culture medium and thereby search for synergistic effects on tissue differentiation. For example, we have added retinoic acid or other morphogens to some cultures (Fig. 2) and observed renal tubule differentiation. We have also tested the combination of activin and bFGF on animal caps. Those factors functioned synergistically for blood cell differentiation (unpublished data). The foregoing procedure describes “direct, single-culture” assays in which animal caps are treated with activin and, after a suitable culturing period, are assayed for type of differentiation. It is also possible, however, to harvest that “direct, single-cultured” tissue and use it as an inducing tissue by combining it with naive (uninduced) tissue in so-called “animal cap combination” assays. Figure 4 illustrates the general procedure. Such combination assays are especially valuable for studying serial inductive interactions, as described in Chapter 11 in Volume I.

2.1.1. Cautionary Notes When preparing to perform a tissue-culture assay or to compare the data an assay has generated with information collected by other laboratories, the following checklist can serve as a guide. Often, data collected in different laboratories has been compared without careful reference to possible differences in the experimental conditions used to collect it, thus these cautions should be carefully considered. _____ (1) Has the activin preparation been biologically calibrated so that doses used by different laboratories can be compared?

20

Asashima et al.

Table 1 Candidate Molecular Indices of Gene Expression that Are Likely Associated with Activin Treatment of Animal Caps Onset of expression

Gene Early-response genes Mix 1 goosecoid Xbra Xlim-1 XFKHI Middle-response genes Xwnt-8 Xnot noggin chordin follistatin cerverus Xhox3 Xprx-1 XMyoD Late-response genes X1Hbox1

30 min 30 min 1.5 h 2h

Relative of Drosophila/mouse oncogene int-1 Homeobox gene, relative of Drosophila empty spiracles BMP-binding protein BMP-binding protein Activin and BMP-binding protein Head inducer. BMP, Wnt, nodal-binding protein Homeobox gene, relative of Drosophila even skipped Drosophila paired related homeobox gene Homolog of mouse MyoD

11 ha

Homeobox gene, relative of Drosophila antennapedia, homologue of mouse Hox C6 Homeobox gene, relative of Drosophila abdominal-B, homologue of mouse Hox B9 Cardiac and early axial muscle actin Neural cell-adhesion molecule gene Homeobox gene, expressed in pancreas

13 ha

a-actin N-CAM X1Hbox8

19 h 19 h 36 ha

_____ _____ _____ _____ _____

(2) (3) (4) (5) (6)

Homeobox gene, relative of Drosophila paired Homeobox gene, relative of Drosophila goosebery and bicoid Homolog of mouse Brachyury (T) Homeobox gene with LIM domain Homeobox gene, relative of Drosophila forkhead

3 ha 3 ha 3 ha 3 ha 3 ha 3 ha 6h* 6h 6 ha

X1Hbox6

aApproximate

Defining feature of gene

time after initiation of activin treatment.

Are similar species of amphibia employed? Are animal caps isolated from similar stage embryos? Are similarly sized animal caps used by different laboratories? Are the animal caps uniform in age* and size? Are the histological criteria or molecular markers used to score differentiation comparable?

2.2. Injection Assays Microinjection of either activin mRNA or activin protein into blastomeres or blastocoel of the early embryo can be carried out without much difficulty if microinjection equipment is available. By injecting into the uncleaved egg, the advantage *Because Xenopus embryos develop so quickly, unless temperature-regulated conditions are employed it is likely that the last batch of animal caps prepared for assay will have developed a significant extent beyond the stage of the first-collected batch.

Activin in Amphibian Embryogenesis

21

Fig. 4. “Animal cap combination” assay in which an ectodermal cap (in this example, from the urodele [newt] Cynops pyrrhogaster) is treated with activin in the tissue culture assay, and precultured for either a short or long period of time, then inserted as a component of a sandwich culture with naive ectoderm and further cultured. In this instance, the activin-treated animal cap mimics the properties of early- and late-stage Spemann primary embryonic organizers.

accrues—in principle—that all cells of the early embryo are exposed to the injected component. Thus, this assay can be employed to measure the effects of activin on patterning processes (e.g., axis formation) in the whole, intact embryo because virtually all cells in the early embryo are exposed to activin. Figure 5 illustrates a typical protocol for a simple injection assay. As well, by dissection of the animal cap from an embryo that previously had been injected with either activin protein (see ref. 16) or activin mRNA and using that animal cap as prospective inducing tissue in the manner illustrated in Fig. 4, the utility of injection assays can be expanded.

22

Asashima et al.

Fig. 5. Microinjection of either activin mRNA or protein into the four-cell stage or the blastocoel of an early embryo has effects on subsequent pattern formation.

2.2.1. Cautionary Notes For the injection of activin protein, many of the same above-mentioned cautions for the tissue-culture assay should be considered. For mRNA injection, yet another set of cautions should be carefully considered: _____ (1) Is the mRNA preparation fully intact, or partially degraded? _____ (2) Is the mRNA actually translated into a functional protein once inside the egg?* _____ (3) Is the mRNA preparation distributed uniformly throughout the egg, or does it remain as a bolus (near the injection site)? _____ (4) Is the mRNA sufficiently long-lived, once injected into an egg, to give a satisfactory result? _____ (5) Is a “no effect” result meaningful?

2.3. Interference–Injection Assays Because it is well known that peptide growth factors such as activin must interact with specific protein receptors in order to exert their ultimate effects on differentiation and morphogenesis, microinjection of either excess amounts of the normal *This criterion is, unfortunately, seldom addressed.

Activin in Amphibian Embryogenesis

23

activin receptor (mRNA) or defective variants of that receptor (mRNA) provides a way to test the role endogenous activin plays in differentiation and morphogenesis. A typical assay protocol would be the same as shown in Fig. 5. For example, Kondo et al. (17) and Hemmati-Brivanlou et al. (18) overexpressed the activin receptor in Xenopus embryos and observed dramatic effects on patterning. Their strategy was to inject copious amounts of the mRNA into fertilized, uncleaved eggs. More recently, Armes and Smith (19) have observed the development of a secondary axis when an activin receptor mRNA was injected into the ventral side of 32-cell stage embryos. Thus, patterning effects can be obtained by the appropriate administration of receptor mRNA. In addition, microinjection of defective variants of the activin receptor mRNA has been performed. Such dominant negative activin receptor mRNA injections have been accomplished by several laboratories. For example, Hemmati-Brivanlou and Melton (20) observed that embryonic cells normally fated to become epidermis developed neural traits when they were injected with a truncated activin receptor mRNA. New et al. (21) discovered that defective receptor mRNA caused axial defects in some circumstances and secondary axes in others. Thus, the use of these bioassay formats designed to interfere with endogenous functions of activin circuitry provide fascinating chances for further experimentation and a plethora of opportunities for speculation regarding the mechanism(s) of action of activin!

2.3.1. Cautionary Notes Similar considerations to those expressed for the injection assays should, of course, be monitored in these interference–injection assays. In addition, the following points are pertinent: _____ (1) Is the presumptive activin receptor mRNA translated into a protein which binds activin in vivo? _____ (2) Does the pattern of accumulation of the activin receptor protein coincide with tissue locations where activin is expected to act? _____ (3) What is the meaning of a “no effect” result?

3. Methods General conditions for operating on amphibian embryos are described in Chapter 11, Volume I. Many of the same methods can be used at the “front end” of the activin bioassays described herein. For example, many of the animal cap assay methods are described in that chapter, as are details and photographs of the microinjection techniques which would be used in the injection bioassays described herein. The details of additional methods are described below. Histological methods vary of course, depending on the size and source of the tissue, as well as the types of observations intended. Our companion chapter in Volume I can be consulted for some protocols. Activin/antiactivin antibody suppliers include Innogenetics, S.A., Ghent, Belgium and Austral Biologicals, San Raman, CA, USA. For marking cells to trace their fate in these bioassays 10–50 nL of 1% Texas Red– dextran–amine (TRDA, D-1863; Molecular Probes, Eugene, OR) or 10–50 nL of 1%

24

Asashima et al.

fluorescein–dextran–amine (FDA, D-1820; Molecular Probes) are injected at the early cleavage stage (e.g., two or four cells) into single blastomeres. 4. Concluding Remarks Activins have been reported to display a variety of effects on many different tissues and cell types at numerous stages in the growth and development of vertebrates. A remarkable array of effects has been reported in a broad range of test systems (reviewed in ref. 22). In no instances, however, is the molecular basis of action fully understood. Activin likely functions in the context of an assemblage of regulatory molecules, including activin receptors and inhibitors (e.g., follistatin), and in combination with other regulatory molecules (e.g., other peptide growth factors). Unraveling its mechanism(s) of action will probably require the use of several different experimental approaches, many of which might include the use of the types of bioassays described in this chapter. The inherent simplicity of the bioassays described herein should be attractive for reductionist approaches to discovering the details associated with activin action. It will certainly be interesting to—ultimately—establish the similarities and differences between activin action in, say, the amphibian egg compared to smooth muscle. These bioassays have the advantage that they can serve as a model for the characterization of not only the effects on differentiation when activin is studied alone, but also the action of combinations of growth factors. In addition to employing these assays for the reductionist-type analyses for which they are especially well suited, it is possible to use them to generate holistic approaches to understanding embryonic patterning. For example, we have embarked on the A-FEAT (Assembling a Functioning Embryo from Activin-Treated components) project, an attempt to answer the following question: What are the minimal morphogen and cell-mass requirements for complete patterning of the tissues and organs in an early embryo?

Restated as a hypothesis: Activin and animal cap cells are sufficient, when manipulated in an appropriate fashion, to construct a whole, functioning embryo!

The observation of the remarkable ability of activin to induce in animal caps a wide range of tissues and organs (e.g., Fig. 2) is providing a foundation for the A-FEAT project, an attempt to determine what components are sufficient for patterning a whole embryo. Activin alone at various concentrations, as well as in combination with other prospective patterning molecules (e.g., other peptide growth factors), will provide a research platform, based largely on the bioassays described in this chapter. Acknowledgments We are grateful to members of our laboratories and our colleagues for useful discussions (e.g., Prof. Anton Neff) and editing (Susan Duhon) of this manuscript. Most of this work was supported by Grants-in-Aid for Scientific Research from the Ministry of Education, Science, Sports and Culture of Japan, and by CREST (Core Research for Evolutional Science and Technology) of the Japan Science and Technology Corporation. G.M.M.’s research is funded by the National Science Foundation (USA).

Activin in Amphibian Embryogenesis

25

References 1. Murata, M., Eto, Y., Shibai, H., Sakai, M., and Muramatsu, M. (1988) Erythroid differentiation factor is encoded by the same mRNA as that of inhibin βA chain. Proc. Natl. Acad. Sci. USA 85, 2434–2438. 2. Tiedemann, H., Asashima, M., Born, J., Grunz, H., Knochel, W., and Tiedemann, H. (1996) Determination, inductin and pattern formation in early amphibian embryos. Dev. Growth Differ. 36, 233–246. 3. Smith, J. C., Price, B. M. J., van Nimmen, K., and Huylebroek, D. (1990) Identification of a potent Xenopus mesoderm inducing factor as a homologue of activin A. Nature 345, 729–731. 4. Asashima, M., Nakano, H., Shimada, K., Kinoshita, K., Ishi, K., Shibai, H., and Ueno, N. (1990) Mesodermal induction in early amphibian embryos by activin A. Roux’s Arch. Dev. Biol. 198, 330–335. 5. Fukui, A., Nakamura, T., Uchiyama, H., Sugino, K., and Asashima, M. (1994) Identification of activins A, AB, and B and follistatin proteins in Xenopus embryos. Dev. Biol. 163, 279–281. 6. Asashima, M., Nakano, H., Uchiyama, H., Sugino, H., Nakamura, T., Eto, Y, Ejima, D., Nishimatsu, S., Ueno, N., and Kinoshita, K. (1991) Presence of activin (erythroid differentiation factor) in unfertilized eggs and blastulae of Xenopus laevis. Proc. Natl. Acad. Sci. USA 88, 6511–6514. 7. Uchiyama, H., Nakamura, T., Komazaki, S., Takio, K., Asashima, M., and Sugino, H. (1994) Localization of activin and follistatin proteins in the Xenopus oocyte. Biochem. Biophys. Res. Commun. 202, 484–489. 8. Dohrmann, C. E., Hemmati-Brivanlou, A., Thomsen, G. H., Fields, A., Woolf, T. M., and Melton, D. A. (1993) Expression of activin mRNA during early development in Xenopus laevis. Dev. Biol. 157, 474–483. 9. Okabayashi, K., Shoji, H., Nakamura, O., Nakamura, T., Hashimoto, O., Asashima, M., and Sugino, H. (1996) cDNA cloning and expression of the Xenopus laevis vitellogenin receptor. Biochem. Biophys. Res. Commun. 224, 406–413. 10. Dohrmann, C. A., Hemmati-Brivanlou, A., Thomsen, G. H., Fields, A., Woolf, T. M., and Melton, D. A. (1993) Expression of activin mRNA during early development of Xenopus laevis. Dev. Biol. 157, 474–483. 11. Ariizumi, T., Moriya, N., Uchiyama, H., and Asashima, M. (1991) Concentration dependent inducing activity of activin A. Roux’s Arch. Dev. Biol. 200, 230–233. 12. Green, J. B. and Smith, J. C. (1990) Graded changes in dose of a Xenopus activin A homologue elicit stepwise transitions in embryonic cell fate. Nature 347, 391–394. 13. Ariizumi, T. and Asashima, M. (1995) Head and trunk-tail organizing effects of the gastrula ectoderm of Cynops pyrrogaster after treatment with activin A. Roux’s Arch. Dev. Biol. 204, 427–435. 14. Uochi, T., and Asashima, M. (1996) Sequential gene expression during pronephric tubule formation in vitro in Xenopus ectoderm. Dev. Growth Differ. 38, 625–634. 15. Ariizumi, T., Komazaki, S., Asashima, M., and Malacinski, G. M. (1996) Activin treated urodele ectoderm: a model experimental system for cardiogenesis. Int. J. Dev. Biol. 40, 715–718. 16. Ariizumi, T., Sawamura, K., Uchiyama, H., and Asashima, M. (1991) Dose and timedependent mesoderm induction and outgrowth formation by activin A in Xenopus laevis. Int. J. Dev. Biol. 35, 407–414. 17. Kondo, M., Tashiro, K., Fujii, G., Asano, M., Miyoshi, R., Yamada, R., Muramatsu, M., and Shiokawa, K. (1991) Activin receptor mRNA is expressed early in Xenopus embryo-

26

18. 19.

20. 21.

22.

Asashima et al. genesis and the level of the expression affects the body axis formation. Biochem. Biophys. Res. Commun. 181, 684–690. Hemmati-Brivanlou, A., Wright, D. A., and Melton, D. A. (1992) Embryonic expression and functional analysis of a Xenopus activin receptor. Dev. Dyn. 194, 1–11. Armes, N. A. and Smith, J. C. (1997) The ALK-2 and ALK-4 activin receptors transduce distinct mesoderm-inducing signals during early Xenopus development but do not cooperate to establish thresholds. Development 124, 3797–3804. Hemmati-Brivanlou, A., and Melton, D. A. (1994) Inhibition of activin receptor signaling promotes neuralization in Xenopus. Cell 77, 273–281. New, H. V., Kavka, A. I., Smith, J. C., and Green, J. B. (1997) Differential effects on Xenopus development of interference with type IIA and type IIB activin receptors. Mech. Dev. 61, 175–186. Ying, S.-Y., Zhang, Z., Furst, B., Batres, Y., Huang, G., and Li, G. (1997) Activins and activin receptors in cell growth. Proc. Soc. Exp. Biol. Med. 214, 114–122.

Morphogenesis Analysis

27

4 Analysis of Mammary Gland Morphogenesis Calvin D. Roskelley, Colleen Wu, and Aruna M. Somasiri 1. Introduction Mammary gland development is marked by numerous cellular changes that begin in the embryo, continue postnatally, are modified at puberty, and come to full fruition during the adult cycles of pregnancy, lactation, and involution. These changes are initiated by hormones produced at distant sites and by the microenvironment within the gland itself. The latter, which includes soluble factors and the insoluble extracellular matrix (ECM), will be the focus of this chapter. In the embryo, mammary gland morphogenesis begins when epidermal protrusions branch within an underlying adipocytic mesenchyme to form the preliminary ductal tree (1). At birth, the epithelial and mesenchymal compartments of the developing gland are separated by a continuous layer of specialized ECM, the basement membrane (2). The deposition of laminin and heparan sulfate proteoglycan into the basement membrane by the adipocytic mesenchyme is crucial for further epithelial development. In males, androgen-mediated conversion to a fibroelastic mesenchyme results in fibronectin and tenascin deposition that sparks ductal regression (2–4). Thus, the specific molecular composition of the basement membrane is an important mediator of early ductal outgrowth. Postnatally, a second phase of morphogenesis occurs as growth and branching of the ductal tree accelerates. Epithelial proliferation at puberty is directed by ovarian hormones that trigger mesenchymal production of locally acting growth factors such as transforming growth factor-α (TGF-α) (5). Other factors produced by the mesenchyme alter ECM-dependent branching decisions. For example, localized production of TGF-β causes the deposition of a fibroelastic ECM that prevents branching along the walls of the ducts (6). In contrast, site-specific side branching is induced by hepatocyte growth factor (HGF; see ref. 7). Despite the fact that HGF is produced by the mesenchyme, its phenotypic actions are mediated by receptors on the surface of epithelial cells (8). Although it is not yet known whether HGF interacts directly with the ECM, it is clear that the growth factor alters interactions between mammary epithelial cell interactions and collagen matrices (9,10). Lobulo-alveolar development takes place during pregnancy. Estrogens, progestins, and prolactin are critical regulators of the massive epithelial proliferation that takes From: Methods in Molecular Biology, Vol. 136: Developmental Biology Protocols, Vol. II Edited by: R. S. Tuan and C. W. Lo © Humana Press Inc., Totowa, NJ

27

28

Roskelley, Wu, and Somasiri

place early in this third morphogenic phase. This short-lived hyperplasia is kinetically coupled to the formation of small alveolar outpocketings that bud from the lateral margins of the terminal portions of the ductal tree. The latter process is regulated, at least in part, by another mesenchymally produced growth factor, neuregulin (7). As alveolar budding proceeds, an intimate association between the epithelial cells and the underlying basement membrane is maintained. This association is absolutely required for further alveolar morphogenesis, milk protein production, and lactational secretion (11). Later, after weaning, the targeted destruction of the basement membrane induces epithelial apoptosis, which is the major force driving glandular involution (12–14). We have long been interested in those interactions between the mammary epithelium and the surrounding basement membrane that regulate alveolar morphogenesis (15,16). Mammary epithelial cells isolated from mid-pregnant mice rapidly dedifferentiate in monolayer culture. However, when they are placed on a reconstituted basement membrane gel, these cells round up, aggregate, form cell–cell junctions, and become polarized. This produces small alveolar “mammospheres” that cavitate to form a central lumen. In an appropriate hormonal milieu, cells within these mammospheres fully differentiate. Specifically, the cells express milk protein genes, package the products in secretory vesicles, transport the vesicles to the apical cell surface, and release the vesicular products vectorially into the central lumen. In an effort to identify the mechanisms responsible for regulating basement-membrane-dependent alveolar morphogenesis, we developed a battery of specialized culture models. First, we isolated a functional mouse mammary epithelial cell line that is not capable of depositing its own endogenous basement membrane. We have used this cell line for the following: 1. Suppress or enhance epithelial-to-mesenchymal transitions (EMT). 2. Manipulate cell shape. 3. Identify the basement membrane ligand and the integrin receptor responsible for milk protein gene expression. 4. Examine integrin-mediated signal transduction. 5. Determine how each of these events contributes to the complete alveolar morphogenesis that occurs during mammosphere formation.

These models are now described in detail (see Fig. 1). 2. Experimental Models and Approaches 2.1. Model 1: Epithelial-to-Mesenchymal Transition

2.1.1. Rationale To examine ECM-dependent differentiation, we isolated a basement-membraneresponsive mouse mammary epithelial cell line by limited dilution cloning (17). This line, designated scp2, is made up of cells that contain keratin, but not vimentin, intermediate filaments. Approximately 90% of these cells respond to a combined treatment with lactogenic hormones and basement membrane by expressing the milk protein β-casein (18). When they are routinely maintained and passaged in a two-dimensional (2-D) monolayer culture, a small proportion of scp2 cells “drift” from an epithelial to a fibroblastic morphology. The latter exhibit decreased keratin expression and increased vimentin

Morphogenesis Analysis

29

Fig. 1. A schematic representation of the five tissue culture models used to study structural and functional events that initiate alveolar morphogenesis by mammary epithelial cells.

expression and cannot be induced to express milk protein genes. In addition, cell–cell interactions are disrupted, and the cells invade basement membrane gels. It is not yet clear whether this phenotypic drift, which resembles an epithelial-to-mesenchymal transition (EMT) (19), indicates that the original scp2 clone had stem-cell characteristics. Regardless, this potential can be used to test the efficacy of suspected EMT-inducing agents or to examine epithelial—mesenchymal interactions in the emerging mixed cultures. However, during routine culture, it is best to minimize this conversion, because it leads to a decreased stringency in the response to exogenously added basement membrane. We have identified three factors that contribute to the phenotypic drift of mammary epithelial cells during routine 2-D culture: high serum concentrations, growth past confluence, and overtrypsinization. Thus, we have developed a protocol to minimize these factors. Additionally, to ensure that our stocks remain epithelial, we use differential adhesion to remove fibroblastic cells at the time of passaging.

2.1.2. Methods 1. Rinse a subconfluent (approx 80%) monolayer culture of mammary epithelial cells (i.e., mouse epithelial clone scp2; see ref. 17) in a 100-mm dish with serum-free Dulbecco’s modified Eagle’s medium (DMEM)/F12 medium, maintain at 37°C for 10 min, and repeat 3×. This removes any adhesive serum factors from the cultures. 2. Rinse with Ca2+/Mg2+-free DMEM/F12 and maintain at 37°C for 10 min. This reduces cell–cell adhesion and dramatically decreases the time required for trypsinization. 3. Add 2 mL of 0.05% trypsin/0.02% EDTA in Hank’s balance salt solution (HBSS) at room temperature and leave until the cells begin to round on the plate. This takes approx 2–5 min.

30

Roskelley, Wu, and Somasiri

4. Remove trypsin, add 2 mL fresh trypsin, maintain at 37°C until all cells detach (5–15 min). Using this low concentration of trypsin (0.05%), the cells often detach as clusters, and, unless a specific experiment requires single cells, this is optimal, as it slows fibroblastic conversion. 5. Add 8 mL of 5% FBS containing DMEM/F12 to quench trypsin; centrifuge at 200g for 5 min. 6. Resuspend in 12-mL of attachment medium. This is 5% fetal bovine serum (FBS) DMEM/F12 containing insulin (5 µg/mL) and gentamycin (50 µg/mL). Insulin at this concentration acts as a growth factor supplement and prevents apoptosis in routine culture. 7. Plate in 1 × 100-mm dish, incubate 30 min at 37°C. Gently remove unattached cells and plate in 4 × 100-mm dishes (3 mL per dish, 1:4 split), add 7 mL of medium to each. The four dishes containing the cells that did not rapidly adhere to the original dish are epithelial. The rapidly adherent cells left behind in the original culture dish have converted to a fibroblastic phenotype. 8. After overnight attachment of the epithelial cell population, change the medium to routine growth medium. This is DMEM/F12 supplemented with 3% newborn calf serum, insulin, and gentamycin. This medium is changed every second day until the cells are once again subconfluent, which takes approx 4–6 d. The maintenance of the epithelial cells in medium containing a low percentage of calf serum is sufficient to ensure robust cell growth (with insulin), but it decreases the fibroblastic conversion observed when fetal serum is used. It is also important not to overgrow the monolayers past confluence to prevent fibroblastic conversion. 9. Fibroblastic cells in the original adherent dish can also be maintained with growth medium and passaged as above except that trypsinization times will be increased.

2.1.3. Experimental Uses of the Model 1. Phenotypic drift occurs in other mammary epithelial lines. Despite the fact that the HC-11 mouse mammary cell line was isolated by limited dilution cloning, it is heterogeneous, and only approx 10–20% of the cells are capable of expressing the milk protein genes (20). As a result of this heterogeneity, when HC-11 cultures are maintained at very high cell densities, they deposit an endogenous basement membrane. Thus, β-casein expression can be induced in the absence of exogenous basement membrane addition (21). This has proven very useful in experiments designed to elucidate functional lactogenic hormone signaling (22). 2. Mixing of epithelial and fibroblastic cultures. Epithelial and fibroblastic mouse mammary cell lines have been intentionally mixed to induce the deposition of endogenously produced extracellular matrix proteins. This was used to demonstrate that the basement membrane glycoprotein laminin is deposited at sites of direct contact between the two cell types (23). 3. Epithelial-to-mesenchymal transition and invasion. EMT plays an important role in tumorigenesis (24). Therefore, we have assayed the ability of various agents to facilitate this transition in scp2 mammary epithelial cells. Two efficient inducers of EMT are an activated form of the metalloprotease stomelysin-1 (25) and the integrin-linked kinase (26). The latter is a signaling molecule that disrupts integrin-mediated cell adhesion and accelerates the mesenchymal conversion of scp2 cells by decreasing E-cadherin expression and translocating the resulting free β-catenin to the nucleus.

2.2. Model 2: Cell-Shape Manipulation 2.2.1. Rationale Mammary epithelial cells form flat, nondescript monolayers that do not express milk proteins after attachment to chemically crosslinked basement membrane matrices (27).

Morphogenesis Analysis

31

A critical differentiative component that is lost after attachment to such rigid matrices is cell rounding. This has been demonstrated by forcing the cells to round in the absence of ECM. Under the latter conditions, proliferation ceases and expression of the ironbinding milk protein lactoferrin is initiated (28). It is important to point out that this mechanical (i.e., mediated by cell shape) induction of differentiation is only partial. Expression of other milk protein genes does not occur. Cell rounding also acts to initiate the differentiation of hepatocytes (29), retinal pigment epithelial cells (30), keratinocytes (31), and steroidogenic cells (32). Cell shape can be manipulated in a number of ways. Adherent cells can be maintained in suspension, which causes them to become spheroidal, although this often initiates apoptotic cell death within hours (33,34). Cells can also be plated on flexible filters. Alternatively, rounded cells can be produced by restricting cell spreading after attachment to a rigid substratum. This has been accomplished using “elastomeric stamps” that produce microscopic adhesive islands of precise area. Such methodology has been used to demonstrate definitively that cell shape alone regulates a shift between proliferative and differentiative phenotypes in hepatocytes (35). Generalized decreases in adhesivity, which allow for cell attachment but not for spreading, have also been used to produce rounded, differentiated keratinocytes (31). We have used the latter approach to culture rounded mammary epithelial cells. To do this, the cells are plated on the inert, nonadhesive substatum poly-hydroxyethyl methacrylate (polyHEMA; see ref. 18). The cells attach to low concentrations of this substratum and form roundedcell clusters. Alternatively, single rounded cells can be produced by lowering the calcium concentration during cell plating to prevent cell–cell interactions (28).

2.2.2. Methods 1. Dissolve the antiadhesive polymer polyHEMA (Sigma Chemical Co., St. Louis, MO, USA) in 95% ethanol at 50 mg/mL. Maintain this stock in a water bath at 37°C overnight to ensure that all polyHEMA is in solution. 2. Dilute the stock polyHEMA solution 1:100 with 95% ethanol (0.5 mg/mL final concentration) and coat regular tissue culture dishes with 125 µL/cm2 surface area (i.e., 1.0 mL/35-mm dish, 2.5 mL/60-mm dish, 5.0 mL/100-mm dish). Evaporate the ethanol overnight in a dry, sterile incubator at 37°C. This will leave a clear polyHEMA coating on the surface of the dish. 3. Plate scp2 (or other mammary epithelial) cells on polyHEMA-coated plates at 5 × 104 cells/cm2 surface area (i.e., 4 × 105/35-mm dish, 1.0 × 106/60-mm dish, 2.0 × 106/100-mm dish) in DMEM/F12 medium containing 1% FBS and gentamycin (50 µg/mL) supplemented with the lactogenic hormones insulin (5 µg/mL), hydrocortisone (1.0 µg/mL), and prolactin (3 µg/mL). 4. Allow cells to attach overnight and then change to serum-free DMEM/F12 containing lactogenic hormones. 5. Lactoferrin expression is initiated within 48 h, and this can be assessed by Northern blotting, immunofluorescence, Western blotting, or immunoprecipitation of metabolically labeled cell proteins (see refs. 18 and 28 for specific methods and reagents).

2.2.3. Experimental Uses of the Model 1. The functional importance of changes to the cytoskeleton can be assessed in response to cell rounding. The loss of actin stress fibers induced by cell rounding or by treatment with the f-actin-disrupting agent cytochalasin D contributes to lactoferrin induction (28).

32

Roskelley, Wu, and Somasiri

2. Cell rounding inhibits proliferative responses to growth factors. Insulin-mediated signal transduction is downregulated in rounded scp2 cells. Specifically, MAP kinase pathways are curtailed, which results in a downregulation of ets and AP1 transcription factor activities. Artificial stimulation of this pathway using an activated form of the ras oncogene prevents lactoferrin expression (C. Roskelley, unpublished observations). 3. Cell rounding transcriptionally activates the lactoferrin promoter. Stable transfection of a reporter gene construct has been used to demonstrate that cell rounding transcriptionally activates a 2.6-kb fragment of the mouse lactoferrin promoter (28).

2.3. Model 3: Integrin Interactions with the Basement Membrane 2.3.1. Rationale When flat, nonfunctional mammary epithelial monolayers on tissue culture plastic are overlaid with a solution of basement membrane matrix, the cells round and cluster in a manner similar to that observed when they are plated on antiadhesive polyHEMA (17). Cells under both conditions also express lactoferrin. However, a crucial differentiative difference in the basement-membrane-overlaid condition is the expression of a second milk protein, β-casein (18). Induction of the latter is specifically initiated by laminin, a glycoprotein component of the basement membrane overlay (36).

2.3.2. Method 1. Plate scp2 (or other mammary epithelial cells) on tissue culture dishes at 2.5 × 104/cm2 surface area (i.e., 2 × 105/35-mm dish, 5 × 106/60-mm dish, 1 × 106/100-mm dish) in DMEM/F12 medium containing 1% FBS supplemented with the lactogenic hormones insulin (5 µg/mL), hydrocortisone (1.0 µg/mL), and prolactin (3 µg/mL). 2. Allow cells to attach overnight and then change to serum-free DMEM/F12 containing lactogenic hormones for 24 h to remove growth factors. This ensures that the cells are not proliferating when the ECM overlay is added. 3. Dilute ECM to the desired concentration in cold DMEM/F12 containing lactogenic hormones. Basement membrane ECM (see ref. 37; Matrigel; Collaborative Research Inc., New Bedford, MA, USA) is thawed on ice and is diluted in the medium to a final total protein concentration of 150 µg/mL (approx 1:100 dilution). For purified laminin overlays, we routinely use a concentration of 50 µg/mL. Concentrations for laminin fragments and other ECM proteins vary (see ref. 36 for specific methods and reagents). 4. Warm the medium containing the ECM proteins to 37°C in a water bath, vortex briefly, mix by pipeting (always use plastic pipets when working with ECM-containing medium), and add the medium to the cells immediately. This constitutes the ECM overlay. Medium change the overlay every 2 d. 5. The cells will round and cluster after 2–3 d in response to the overlay. Lactoferrin and β-casein induction occurs soon thereafter, usually between 3 and 5 d. Expression of both milk proteins can be assessed by Northern blotting, immunofluorescence, Western blotting, or immunoprecipitation of metabolically labeled cell proteins (see refs. 18 and 28 for specific methods and reagents).

2.3.3. Experimental Uses of the Model 1. A specific domain of laminin is involved in β-casein induction. Overlays with purified ECM proteins were used to determine that laminin is the specific basement ligand that induces cell rounding and β-casein expression (36). A specific fragment of laminin, derived from the E3 domain, is able to inhibit overlay-mediated β-casein induction.

Morphogenesis Analysis

33

2. β-Casein promoter activation. A basement-membrane-responsive 160-bp element derived from the bovine β-casein promoter (38) is activated by the laminin overlay in this assay, as is a 300-bp fragment from the rat β-casein promoter (39). These elements must be stably integrated into the mammary epithelial cell genome to be ECM responsive. In transient transfection assays, they are only lactogenic hormone responsive (40). 3. More than one cell-surface receptor mediates the differentiative effects of a laminin overlay. Cell-rounding-mediated lactoferrin expression does not require integrin signaling (28), whereas β-casein expression is integrin dependent (41). Therefore, the same basement membrane glycoprotein induces the differentiative expression of two tissue-specific genes using at least two different receptors and/or signaling pathways.

2.4. Model 4: Integrin-Mediated Signaling 2.4.1. Rationale Cell clustering and scp2 mammary epithelial cell rounding are necessary—but on their own, these morphological changes are not sufficient—to induce the β-casein gene expression (18). A second integrin-mediated signal is also required. This signal can be assayed in isolation by overlaying preclustered cells with laminin. Under the latter conditions, β-casein expression is induced very quickly, within hours (16). This short time frame makes it possible to identify specific transduction events that contribute to β-casein induction. This is an important and novel model system for assessing differentiative integrin signaling, as the great majority of other models focus on ECMdependent cell attachment, spreading, migration, and/or proliferation.

2.4.2. Method 1. Plate scp2 (or other mammary epithelial) cells on polyHEMA-coated plates (see Subheading 2.2.2.) at 5 × 104/cm2 surface area (i.e., 4 × 105/35-mm dish, 1.0 × 106/60-mm dish, 2.0 × 106/100-mm dish) in DMEM/F12 medium containing 1% FBS supplemented with lactogenic hormones insulin (5 µg/mL), hydrocortisone (1.0 µg/mL), and prolactin (3 µg/mL). 2. Allow cells to attach overnight and then change to serum-free DMEM/F12 containing lactogenic hormones; leave for 48 h. These “naked” cell clusters express lactoferrin but not β-casein. 3. Dilute laminin in cold, serum-free DMEM/F12 medium containing lactogenic hormones (see step 1). Warm medium to 37°C and add this overlay to the preclustered cells immediately. There is no further overt morphological change, but β-casein expression is rapidly initiated.

2.4.3. Experimental Uses of the Model 1. Cell rounding “primes” the cells to rapidly respond to laminin: Once they are rounded and clustered (i.e., on polyHEMA), a laminin overlay initiates rapid induction of β-casein expression within 4–8 h (16,18). This rapid induction is regulated, at least in part, transcriptionally, when a 300-bp fragment of the rat β-casein promoter is activated using this protocol (C. Roskelley, unpublished observations). 2. Integrin-mediated tyrosine phosphorylation contributes to β-casein expression. The laminin overlay initiates an increase in tyrosine phosphorylation in at least seven protein species (18), a signaling event that is required for efficient β-casein induction (42). 3. Activation of the focal adhesion kinase. The laminin overlay causes increased phosphorylation of the focal adhesion kinase, which is an indication of functional activation (16).

34

Roskelley, Wu, and Somasiri

4. Tight-junction formation. Tight junctions are found in laminin-overlaid cell clusters. If these junctions are disrupted, β-casein induction does not occur (A. Somasiri and C. Roskelley, unpublished observations).

2.5. Model 5: Complete Alveolar Morphogenesis 2.5.1. Rationale In vivo, lactoferrin and β-casein expression are initiated during early and mid-pregnancy, respectively. At these developmental stages, alveolar morphogenesis is not complete and the mammary epithelium is not lactational. In culture, laminin-overlaid cell clusters expressing lactoferrin and β-casein are also not lactational. There is no central lumen and no vectorial secretion. The latter only occurs when the cells are cultured as alveolus-like mammospheres on reconstituted basement membrane gels. Alveolar morphogenesis in culture also induces whey acidic protein (WAP) production, a third milk protein gene that is first expressed just prior to lactation in vivo. Prior to mammosphere formation, the basement-membrane gel causes mammary epithelial cells to round up, aggregate, and form cell–cell junctions. Each of these morphogenic events has been examined in isolation using the models in Subheadings 2.1.–2.4. However, additional morphogenic changes must occur prior to lactational WAP expression. These include the acquisition of apical basal polarity, endogenous basement-membrane deposition, and cavitation of a central lumen.

2.5.2. Method 1. Thaw basement-membrane ECM (Matrigel, Collaborative Research) on ice. Place tissue culture dishes on ice in a laminar flow hood. Spread cold, undiluted liquid Matrigel on the bottom of the dish (200 µL/35-mm dish, 500 µL/60-mm dish, 1 mL/100-mm dish). Place dishes in a 37°C humidified incubator for 1 h prior to adding cells. This produces a malleable basement-membrane gel on the bottom of the dish. 2. Plate scp2 (or other mammary epithelial cells) on Matrigel-coated plates at 5 × 104/cm2 surface area (i.e., 4 × 105/35-mm dish, 1.0 × 106/60-mm dish, 2.0 × 106/100-mm dish) in DMEM/F12 medium containing 1% FBS supplemented with the lactogenic hormones insulin (5 µg/mL), hydrocortisone (1 µg/mL), and prolactin (3 µg/mL). The cells attach to the gel and begin to form clusters almost immediately. 3. Allow cells to attach overnight and then change to serum-free DMEM/F12 containing lactogenic hormones. This medium should be changed every second day. The cells begin to express milk proteins after 2–3 d. Polarized “mammospheres,” which secrete milk proteins vectorially into a central lumen, form after 5–7 d.

2.5.3. Experimental Uses of the Model 1. Downregulation of growth factor production. As mammospheres form, the expression of autocrine growth factors by the mammary epithelial cells is downregulated. The growth factors affected include TGF-α (43), which causes alveolar hyperplasia when force-expressed in transgenic mice (44), and TGF-β (45), which inhibits ductal branching during postpubertal development (6). 2. Disruption of integrin signaling initiates apoptosis. Weaning triggers a massive apoptosis of the lactational mammary epithelium that drives alveolar involution in vivo (14). This process is initiated by a metalloprotease-mediated destruction of the basement membrane (12,13,46). Forced overexpression of an activated metalloprotease or function-blocking integrin antibodies both initiate apoptosis in the cells of cultured mammospheres (47,48).

Morphogenesis Analysis

35

3. Distinguishing between normal and tumor-derived mammary epithelial cells. In (2-D) monolayer culture, it is very difficult to distinguish normal from tumor-derived human mammary epithelial cells. However, when they are cultured within basement membrane gels, normal cells form polarized 3-D mammospheres, whereas tumor-derived cells form loose, disorganized proliferating masses that do not exhibit apical–basal polarity (49). This differential response to the basement membrane is mediated by altered integrin expression and signaling (50). Similarly, forced-expression of oncogenes causes morphogenic changes in 3-D culture that are integrin dependent (51).

3. Conclusion Mammary gland development is directed by long-range-acting hormones and microenvironmental factors within the gland itself. One component of the latter, the basement membrane, plays a key role in alveolar morphogenesis throughout developmental cycles of pregnancy, lactation, and involution. In culture, cellular interactions with the basement membrane induce the formation of “mammospheres” which resemble lactational alveoli in vivo. We have developed five specialized culture models to examine individual aspects of this ECM-dependent process, and experimental manipulation of each model has been used to identify many of the mechanisms responsible for regulating alveolar morphogenesis. Not surprisingly, the co-optation of these regulators plays a significant role in the emergence of a number of functionally and structurally important tumor-associated phenotypes. These include uncontrolled proliferation, disruption of cell–cell junctions, loss of epithelial polarity, and increased invasiveness. Acknowledgments Work in the authors’ laboratory is supported by the Canadian Breast Cancer Research Initiative, the National Cancer Institute of Canada, and the British Columbia Health Research Foundation. References 1. Sakakura, T. (1983) Epithelial–mesenchymal interactions in mammary gland development and its perturbation in relation to tumorigenesis, in Understanding Breast Cancer (Rich, M. A., Hager, J. C., and Furmanski, P., eds.), Marcel Dekker, New York, pp. 261–284. 2. Kimata, K., Sakakura, T., Inaguma, K., Kato, M., and Nishizuka, Y. (1985) Participation of two different mesenchymes in the developing mouse mammary gland: synthesis of basement membrane components by fat pad precursor cells. J. Embryol. Exp. Morphol. 89, 243–257. 3. Hueberger, B., Fitzka, I., Wasner, G., and Kratochwil, K. (1982) Induction of androgen receptor formation by epithelium–mesenchyme interaction in embryonic mouse mammary gland. Proc. Natl. Acad. Sci. USA 79, 2957–2961. 4. Chiquet-Ehrismann, R., Mackie, E. J., Pearson, C. A., and Sakakura, T. (1986) Tenascin: an extracellular matrix protein involved in tissue interactions during fetal development and carcinogenesis. Cell 47, 131–139. 5. Snedeker, S. M., Brown, C. F., and DiAugustine, R. P. (1991) Expression and functional properties of transforming growth factor α and epidermal growth factor during mouse mammary gland ductal morphogenesis. Proc. Natl. Acad. Sci. USA 88, 276–280. 6. Silberstein, G. B., Flanders, K. C., Roberts, A. B., and Daniel, C. W. (1992) Regulation of mammary morphogenesis: evidence for extracellular matrix-mediated inhibition of ductal budding by transforming growth factor-beta 1. Dev. Biol. 152, 354–362.

36

Roskelley, Wu, and Somasiri

7. Yang, Y., Spitzer, E., Meyer, D., Sachs, M., Niemann, C., Hartmann, G., Weidner, K. M., Birchmeier, C., and Birchmeier, W. (1995) Sequential requirement of hepatocyte growth factor and neuregulin in morphogenesis and differentiation of the mammary gland. J. Cell Biol. 131, 215–226. 8. Brinkmann, V., Foroutan, H., Sachs, M., Weidner, K. M., and Birchmeier, W. (1995) Hepatocyte growth factor/scatter factor induces a variety of tissue-specific morphogenic programs in epithelial cells. J. Cell Biol. 131, 1573–1586. 9. Niranjan, B., Buluwela, L., Yant, J., Perusinghe, N., Atherton, A., Phippard, D., Dala, T., Gusterson, B., and Kamalati, T. (1995) HGF/SF: a potent cytokine for mammary growth, morphogenesis and development. Development 121, 2897–2908. 10. Soriano, J. V., Pepper, M. S., Nakamura, T., Orci, L., and Montesano, R. (1995) Hepatocyte growth factor stimulates extensive development of branching duct-like structures by cloned mammary gland epithelial cells. J. Cell Sci. 108, 413–431. 11. Howlett, A. R. and Bissell, M. J. (1993) The influence of tissue microenvironment on the development and function of mammary epithelium. Epithelial Cell Biol. 2, 79–89. 12. Talhouk, R. S., Chin, J. R., Unemori, E. N., Werb, Z., and Bissell, M. J. (1991) Proteinases of the mammary gland: developmental regulation in vivo and vectorial secretion in culture. Development 112, 439–449. 13. Talhouk, R. S., Bissell, M. J., and Werb, Z. (1992) Coordinated expression of extracellular matrix-degrading proteinases and their inhibitors regulates mammary epithelial function during involution. J. Cell Biol. 118, 1271–1282. 14. Strange, R., Li, F., Saurer, S., Burkhardt, A., and Friis, A. A. (1992) Apoptotic cell death and tissue remodelling during mouse mammary gland involution. Development 115, 49–58. 15. Bissell, M. J., Hall, G. J., and Parry, G. (1982) How does the extracellular matrix direct gene expression? J. Theoret. Biol. 99, 31–68. 16. Roskelley, C. D. and Bissell, M. J. (1995) Dynamic reciprocity revisited: a continuous, bidirectional flow of information between cells and the extracellular matrix regulates mammary epithelial cell function. Biochem. Cell Biol. 73, 391–397. 17. Desprez, P. Y., Roskelley, C. D., and Bissell, M. J. (1993) Isolation of functional cell lines from a mouse mammary epithelial cell strain: the importance of basement membrane and cell–cell interactions. Mol. Cell. Differ. 1, 99–110. 18. Roskelley, C. D., Desprez, P. Y., and Bissell, M. J. (1994) Extracellular matrix-dependent tissue-specific gene expression in mammary epithelial cells requires both physical and biochemical signal transduction. Proc. Natl. Acad. Sci. USA 91, 12,378–12,382. 19. Hay, E. D. (1995) An overview of epithelio-mesenchymal transformation. Acta Anatomica 154, 8–20. 20. Ball, R. K., Friis, R. R., Schoenenberger, C. A., Doppler, W., and Groner, B. (1988) Prolactin regulation of B-casein gene expression and of a cytosolic 120-kd protein in a cloned mouse mammary epithelial cell line. EMBO J. 7, 2089–2095. 21. Chammas, R., Taverna, D., Cell, N., Santos, C., and Hynes, N. E. (1994) Laminin and tenascin assembly regulate HC-11 mouse mammary cell differentiation. J. Cell. Sci. 107, 1031–1040. 22. Wortmann, M., Cella, N., Hofer, P., Groner, B., Lieu, X., Heninghausen, L., and Hynes, N. E. (1996) Lactogenic hormone activation of Stat5 and transcription of the β-casein gene in mammary epithelial cells is independent of p42 ERK2 mitogen-activated protein kinase activity. J. Biol. Chem. 271, 31,863. 23. Reichmann, E., Ball, R., Groner, B., and Friss, R. R. (1989) New mammary epithelial and fibroblastic cell clones in co-culture form structures competent to differentiate functionally. J. Cell Biol. 108, 1127–1138.

Morphogenesis Analysis

37

24. Birchmeier, C., Birchmeier, W., and Brand-Sabri, B. (1996) Epithelial-mesenchymal transitions in cancer progression. Acta Anatomica 156, 217–226. 25. Lochter, A., Galsosy, S., Muschler, J., Freedman, N., Werb, Z., and Bissell, M. J. (1997) Matrix metalloproteinase stromelysin-1 triggers a cascade of molecular alterations that leads to stable epithelial-mesenchymal conversion and pre-malignant phenotype in mammary epithelial cells. J. Cell Biol. 139, 1861–1872. 26. Novak, A., Hsu, S., Leung-Hagesteijn, C., Radeva, G., Papkoff, J., Montesano, R., Roskelley, C., Grosschedl, R., and Dedhar, S. (1998) Cell adhesion and the integrin linked kinase (ILK) regulate the LEF-1 and B-catenin signalling pathways. Proc. Natl. Acad. Sci. USA 95, 4374–4379. 27. Li, M., Aggeler, J., Farson, D. A., Hatier, C., Hassell, J., and Bissell, M. J. (1987) Influence of reconstituted basement membrane and its components on casein gene expression and secretion. Proc. Natl. Acad. Sci. USA 82, 1419–1493. 28. Close, M. J., Howlett, A. R., Roskelley, C. D., Desprez, P. Y., Bailey, N., Rowning, B., Teng, C. T., Stampfer, M. R., and Yaswen, P. (1997) Lactoferrin expression in mammary epithelial cells is mediated by changes in cell shape and actin cytoskeleton. J. Cell Sci. 110, 2861–2871. 29. Mooney, D., Hansen, L., Vacanti, J., Langer, R., Farmer, S., and Ingber, D. (1992) Switching from differentiation to growth in hepatocytes: control by extracellular matrix. J. Cell. Physiol. 151, 497–505. 30. Opas, M. (1989) Expression of the differentiated phenotype by epithelial cells is regulated by both biochemistry and mechanics of substratum. Dev. Biol. 131, 281–293. 31. Watt, F., Jordan, P. W., and O’Neill, C. H. (1988) Cell shape controls terminal differentiation of human epidermal keratinocytes. Proc. Natl. Acad. Sci. USA 85, 5576–5588. 32. Roskelley, C. D. and Auersperg, N. (1993) Mixed parenchymal-stromal populations of rat adrenocortical cells support the proliferation and differentiation of steroidogenic cells. Differentiation 55, 37–45. 33. Meredith, J. E., Fazeli, B., and Schwartz, M. A. (1993) The extracellular matrix as a cell survival factor. Mol. Biol. Cell 4, 953–961 34. Frisch, S. M. and Francis, H. (1994) Disruption of epithelial cell–cell interactions induces apoptosis. J. Cell Biol. 124, 619–626. 35. Wang, N, Butler, J. P., and Ingber, D. E. (1993) Mechanotransduction across the cell surface and through the cytoskeleton. Science 260, 1124. 36. Streuli, C. H., Schmidhauser, C., Bailey, N., Yurchenco, P., Skubitz, P. N., Roskelley, C. D., and Bissell, M. J. (1995) Laminin mediates tissue-specific gene expression in mammary epithelia. J. Cell Biol. 129, 591–603. 37. Kleinman, H. K., McGarvey, M. L., Hassell, J. R., Star, V. L., Cannon, F. B., Laurie, G. W., and Martin, G. R. (1986) Basement membrane complexes with biological activity. Biochemistry 25, 312–318. 38. Schmidhauser, C., Bissell, M. J., Myers, C. A., and Casperson, G. F. (1990) Extracellular matrix and hormones transcriptionally regulate bovine β-casein 5' sequences in stably transfected mouse mammary cells. Proc. Natl. Acad. Sci. USA 87, 9118–9122. 39. Doppler, W., Welte, T., and Phillipp, S. (1995) CCAAT/enhancer-binding protein isoforms are expressed in mammary epithelial cells and bind to multiple sites in the β-casein gene promoter. J. Biol. Chem. 270, 17,962–17,969. 40. Myers, C. A., Schmidhauser, C., Mellintin-Michelotti, J., Fragoso, G., Roskelley, C. D., Casperson, G., Mossi, R., Pujuguet, P., Hager, G., and Bissell, M. J. (1998) Characterization of BCE-1: a transcriptional enhancer regulated by prolactin and extracellu-

38

41.

42. 43.

44. 45. 46.

47. 48.

49.

50.

51.

Roskelley, Wu, and Somasiri lar matrix and modulated by the state of histone acetylation. Mol. Cell Biol. 18, 2184–2195. Streuli, C. H., Bailey, N., and Bissell, M. J. (1991) Control of mammary epithelial differentiation: the separate roles of cell–substratum and cell–cell interaction. J. Cell Biol. 115, 1383–1385. Roskelley, C. D., Srebrow, A., and Bissell, M. J. (1995) A hierarchy of ECM-mediated signalling regulates tissue-specific gene expression. Curr. Opin. Cell Biol. 7, 736–747. Lin, C. Q., Dempsey, P. J., Coffey, R. J., and Bissell, M. J. (1995) Extracellular matrix regulates whey acidic protein gene expression by suppression of TGF-alpha in mouse mammary epithelial cells—Studies in culture and in transgenic mice. J. Cell Biol. 129, 1115–1125. Matsui, Y., Halter, S. A., Holt, J. T., Hogan, B. L., and Coffey, R. J. (1990) Development of mammary hyperplasia and neoplasia in MMTV-TGF alpha transgenic mice. Cell 61, 1147. Streuli, C. H., Schmidhauser, C., Kobrin, M., Bissell, M. J., and Derynck, R. (1993) Extracellular matrix regulates expression of the TGFβ1 gene. J. Cell Biol. 120, 253–260. Sympson, C. J., Talhouk, R. S., Alexander, C. M., Chin, S. K., Clift, S. M., Bissell, M. J., and Werb, Z. (1994) Targeted expression of stromelysin-1 in mammary gland provides evidence for a role of proteinases in branching morphogenesis and the requirement for an intact basement membrane for tissue-specific gene expression. J. Cell Biol. 125, 681–692. Boudreau, N., Sympson, C. J., Werb, Z., and Bissell, M. J. (1995) Suppression of ICE and apoptosis in mammary epithelial cells by extracellular matrix. Science 267, 891–895. Boudreau, N., Werb, Z., and Bissell, M. J. (1996) Suppression of apoptosis by basement membrane requires three-dimensional tissue organization and withdrawal from the cell cycle. Proc. Natl. Acad. Sci. USA 93, 3509–3513. Peterson, O. W., Ronnov-Jessen, L., Howlett, A. R., and Bissell, M. J. (1992) Interaction with basement membrane serves to rapidly distinguish growth and differentiation pattern of normal and malignant human breast epithelial cells. Proc. Natl. Acad. Sci. USA 89, 9064–9068. Weaver, V. M., Peterson, O. W., Wang, F., Larabell, C. A., Briand, P., Damsky, C., and Bissell, M. J. (1997) Reversion of the malignant phenotype of human breast cells in threedimensional culture and in vivo by integrin blocking antibodies. J. Cell Biol. 137, 231–245. Berdichevsky, F. Wetzels, R. Shearer, M., Maringnone, S., Ramaekers, F. C. S., and Taylor-Papdimitriou, J. (1994) Integrin expression in relation to cell phenotype and malignant change in the human breast. Mol. Cell Differ. 2, 255–274.

Cardiac Mesenchyma and Heart Morphogenesis

39

5 Specification of Cardiac Mesenchyme and Heart Morphogenesis In Vitro H. Joseph Yost 1. Introduction One of the key issues in developmental biology is the question of specification. Specification is the commitment of a group of cells to proceed down a developmental pathway without further extraneous signals. In order identify the development period in which cells have acquired all the necessary information to become committed to a developmental pathway, such as the formation of an organ or tissue, cells are explanted from embryos at various stages and cultured in vitro in a neutral medium. Analysis of the tissues that are formed by the explants allows the investigator to define the developmental period during which specification of a tissue is completed. From these results, one can infer that if cell-to-cell signals are involved in specification of the tissue, they must have occurred earlier than the defined developmental period. It is important to clarify the distinctions between analysis of specification by explantation and analysis of cell fate by fate-mapping. Fate-mapping identifies, at a specific stage in development, the location of cells that will contribute to an organ or tissue at a later stage in development. Fate-mapping does not identify the developmental period during which cells acquire a specific developmental identity nor does it identify the mechanisms by which cell identities are conferred. In contrast, explant analysis identifies the extent to which cells are committed to a developmental pathway at the time of their explantation. Because cells might be exposed to additional signals at subsequent stages of development, explant analysis does not necessarily define the developmental fate of the cells if they were left in an unperturbed embryo. The explantation of cells from Xenopus embryos in order to assess their ability to form cardiac structures in vitro has a long history and has allowed investigators to identify early inductive interactions that lead to cardiac mesoderm formation (1,2), developmental restriction of the cardiogenic field (3), regulation of looping morphogenesis (4), and specification of cardiac left–right orientation (5). Early Xenopus embryonic cells contain internal yolk platelets that provide nutrition, which allows explants to be cultured in minimal culture, avoiding the confounding effects of external supplements. In addition, the robust formation of a looped, beating cardiac tube from a From: Methods in Molecular Biology, Vol. 136: Developmental Biology Protocols, Vol. II Edited by: R. S. Tuan and C. W. Lo © Humana Press Inc., Totowa, NJ

39

40

Yost

sheet of mesoderm cultured in vitro serves as a valuable teaching tool and a dramatic example of morphogenesis for students of developmental biology. 2. Materials 1. Culture media: Embryos and explants are grown in 1/3 × Marc’s Modified Ringers solution (MMR), referred to hereafter as R/3. 1× MMR is 0.1M NaCl, 1.8 mM KCl, 2 mM CaCl2, 1 mM MgCl2, and 5 mM HEPES, pH 7.8. A 40 × stock salt solution is prepared without HEPES and pH adjustment, filtered, and stored. The 40 × stock is diluted to 1 × strength, HEPES is added (1.2 g/L) and the 1 × stock is brought to pH 7.8. This 1 × MMR can be stored at 4°C and is diluted to 1/3 × to make R/3. For explant culture, 50 µg/mL gentamicin sulfate is added to R/3. 2. Fixative: MEMFA is 0.1M MOPS, pH 7.4, 2 mM EGTA, 1 mM MgSO4, and 3.7% formaldehyde. A 5 × stock without formaldehyde is filtered and stored for extended periods at 4°C, protected from light. Stock is diluted and formaldehyde is added on the day of use. 3. Forceps: The leading tip of dissecting forceps (Dumont #5, Roboz Surgical, Rockville MD) is beveled at 45° by gently grinding against a flat diamond-coated knife stone. The angle presents a broad surface at the tip to the forceps, which facilitates removal of the vitelline envelope without puncturing the embryo. 4. Tungsten wire knives: Handles are made from 9-in Pasteur pipets. The glass pipets are heated in a flame near the middle of the narrow half of the pipet, drawn out, and snapped off to form a narrow tip at the end of the handle. A 2-cm length of tungsten wire (size 0.001 Ω/ft; California Fine Wire Company, Grover City, CA) is inserted approximately 1 cm into the tip and sealed by heating briefly over a flame. Heating should be brief, sufficient to melt the glass tip around the tungsten wire but not to sublimate the wire. Several knives should be made in advance. Immediately before use, the wire is trimmed with fine scissors so that approximately 0.3 cm extends from the glass handle. 5. Agarose dishes: Dissections are performed in 60 × 15-mm Petri dishes coated on the bottom with agarose. Agarose (1 g, medium grade) in 100 mL water is melted and poured to the level of one-third of the dish; this amount is sufficient to coat the bottoms of approximately 40 dishes. After solidifying, the agarose dishes can be returned to the plastic sleeve (or wrapped in parafilm or plastic wrap) to prevent dehydration and stored at 4°C for up to several months. Before use, dishes are returned to room temperature and filled with R/3. 6. Dissecting microscopes: A good dissecting binocular microscope with a wide range of magnifications (from 6 × to 50×) and fiber light illumination from the sides greatly facilitates embryo and explant manipulation. A stage that can be cooled to 16°C by recirculating water serves to slow embryo development, allowing a longer period for dissections within the desired embryo stages. 7. Protease. Protease (Qiagen, Valencia, CA) is resuspended to 20 mg/mL in R/3 and stored at 4°C.

3. Methods 1. Preparation of embryos: Xenopus laevis females are injected in the dorsal lymph sac with 50 units of pregnant mare’s serum gonadotropin (Sigma) 2 d before egg laying and 800 units of human chorionic gonadotropin (Sigma) 12–16 h before egg laying. Injected females are kept at 16°C overnight. In the morning, eggs are squeezed from the ovulating female into a Petri dish and immediately fertilized with a minced testis suspension in R/3. 2. Removal of jelly coat: The jelly coat must be completely removed to make the embryos accessible to dissection. Embryos are dejellied by pouring them into an Erlenmeyer flask, decanting the R/3, adding 2% cysteine (pH 8.0, in water), and swirling until embryos

Cardiac Mesenchyma and Heart Morphogenesis

41

rapidly settle to the bottom. Then, the cysteine solution is decanted, embryos are rinsed three times by swirling in and decanting R/3. Embryos in R/3 are then poured into a clean Petri dish. The jelly coat can be removed at any time after 10 min of fertilization. Embryos tend to be healthier if the jelly coat is left intact for the first few cell division cycles, and then removed so that healthy embryos can be separated from unfertilized eggs or abnormal embryos. Embryos are raised at temperatures ranging from 16°C to 22°C, which allows targeted stages to be obtained at convenient times. Embryos are staged according to Nieuwkoop and Faber (6). 3. Removal of vitelline membrane (see Note 1): Place embryos at appropriate stage in agar dishes filled with R/3. Remove the vitelline by pinching the top of the embryo at the surface, allowing the ground angle of the forceps to be tangential to the sphere of the embryo. With the second forceps, pinch the vitelline close to the first forceps, and pull the two forceps away from each other. This should release the embryo from the vitelline. For some investigators and some batches of embryos, it is easier to remove embryos without damage at some stages (early blastula and tailbud) than others (gastrula and early neurula). If this is the case, vitellines can be removed early and embryos can be incubated in R/3 in agarose-coated dishes until the appropriate developmental stage is obtained. 4. Microdissection of precardiac mesoderm: As indicated in Fig. 1A, a series of cuts through the embryo with tungsten wire knives and/or the tips of forceps quickly releases the region of the embryo containing precardiac mesoderm, which is subsequently separated from surrounding tissues. First, place the embryo on its side. Cuts 1 and 2 can be done with the wire knife or simultaneously with the tips of the forceps. With practice, the tips of the forceps can be spaced (and measured through a reticule in the microscope eyepiece) so that the size of the explants are consistent from embryo to embryo. Cut 3 releases a block of tissue, the anterior ventral region of the embryo, which contains an outer ectoderm layer, a middle layer that consists of precardiac mesoderm, and a deep layer of endoderm. The positioning of cut 3 will determine the extent of lateral plate mesoderm included in the explant. Examination of Nkx2.5 expression patterns by in situ hybridization in wholemount embryos (7) is useful to align the boundaries of the cuts with the region of cardiogenic mesoderm for the stages under consideration. In open neural plate stages (e.g., N&F stage 15), the precardiac mesoderm is a single layer of translucent mesoderm cells which tightly adheres to the lightly pigmented ectoderm layer. The deep, opaque, yolk-filled endoderm layer can be peeled away from the mesoderm without disturbing in vitro cardiac morphogenesis (5). To remove endoderm from the explant, begin by peeling the deeper layer at the edge of cut 2, the posterior end of the explant. The layers will readily separate toward the anterior edge (near cut 1), at which the endoderm is more tightly adhered to the ectoderm and mesoderm. This junction can be cut away with a tungsten knife. 5. Culturing cardiac explants (see Note 2): The edges of the ectoderm begin to curl inward immediately after explantation. The ventral midline of the explant appears to flex (and can be assisted with forceps) so that the lateral edges begin to join; that is, the edge of cut 3 on the left joins with the edge of cut 3 on the right side. After approximately 30 min, the edges have sealed together, forming an oblong explant in which general anterior–posterior relationships are maintained. Loose cells and debris are removed from the vicinity with a Pasteur pipet. Explants are maintained in R/3 in agarose dishes until the edges of the explants have sealed. After this, the explants can be transferred with a Pasteur pipet to R/3 in dishes without an agarose base. It is convenient to maintain up to four explants in each well of a covered 24-well culture dish. Explants can be cultured at temperatures between 16°C and 22°C. Control embryos, siblings of the embryos from which explants

42

Yost

Fig. 1. Explantation of cardiogenic region from stage 18 (open neural plate) Xenopus embryos. (A) Diagram depicts left lateral view of an embryo at stage 18, with dorsal at the top, anterior (An) to the left, and posterior (P) to the right. The anterior neural plate is indicated by an asterisk. Cardiogenic region is removed by cuts described in the text and depicted as numbers 1 to 3. (B,C) Photomicrographs of explants cultured to sibling embryo stage 45, fixed and stained with MF20 antibody, in which the left–right orientation of cardiac looping is normal (B) or inverted (C). The atrium (a), conus or outflow (c), and ventricle (v) are indicated. Scale bar represents 50 µm. Adapted with permission from (ref. 5). were derived, are maintained in parallel to assess normal cardiac development in vivo. Usually, explants are cultured until control embryos are between stages 42 and 45. By these stages, the outer ectoderm layer of the explant has become transparent, allowing direct observation of cardiac morphogenesis in the explant. 6. Analysis of in vitro cardiac development: Once the outer layer of the explant (ectoderm) has cleared, cardiac morphogenesis can be directly observed. The direction of rhythmic pulsing can be traced by direct observation and by videomicroscopy (4). The left–right orientation of cardiac tube looping can be determined by comparison of the orientation of the three chambers (Fig. 1B,C) in living explants (5). Explants are fixed in MEMFA for 2 h and stepped through an series of rinses of increasing ethanol concentrations, and they can

Cardiac Mesenchyma and Heart Morphogenesis

43

be stored in 100% methanol at –20°C for several months. Explants can be analyzed by immunohistochemistry or by RNA in situ hybridization by standard methods (8,9). Stained explants are cleared in benzyl benzoate : benzyl alcohol (2:1) for photography.

4. Notes 1. Many find that the most difficult part of the explant protocol is removal of the vitelline membrane. An alternative to removal of the vitelline membrane by forceps is to loosen the vitelline membrane by light treatment with protease. Place 1.9 mL R/3 and 0.1 mL of 20 mg/mL protease in a 2-mL Eppendorf flat-bottom tube. Add approximately 30 embryos and cap the tube so that air bubbles are not trapped. Put the tube on end-over-end rotator for 2–4 min, checking embryos intermittently. Remove embryos to an agarose-coated dish and remove loosened vitelline membranes with forceps. Protease solution can be stored at 4°C for subsequent use. If protease is used to remove the vitelline membrane, it would be best do so well before the stage of explantation, so that protease can be rinsed away. 2. During transfers and culturing, explants and embryos without vitelline membranes must not contact the medium meniscus, as surface tension will lyse cells. In transfer pipets, draw up some medium before drawing up embryos or explants, and insert the tip of the pipet into medium before expelling embryos or explants from pipet.

Acknowledgment I thank Dr. Maria Danos for participation in development and refinement of this protocol and Dr. Jeff Essner for discussion of protease treatment and assistance with the figure. This work is supported in part by a grant from the National Institutes of Health (R01 HL57840) and an American Heart Association Established Investigator award. References 1. Sater, A. K. and Jacobson, A. G. (1990) The role of the dorsal lip in the induction of heart mesoderm in Xenopus laevis. Development 108, 461–470. 2. Nascone, N. and Mercola, M. (1995) An inductive role for the endoderm in Xenopus cardiogenesis. Development 121, 515–523. 3. Sater, A. K. and Jacobson, A. G. (1990) The restriction of the heart morphogenetic field in Xenopus laevis. Dev. Biol. 140, 328–336. 4. Yost, H. J. (1990) Inhibition of proteoglycan synthesis eliminates left–right asymmetry in Xenopus laevis cardiac looping. Development 110, 865–874. 5. Danos, M. C. and Yost, H. J. (1996) Role of notochord in specification of cardiac left– right orientation in zebrafish and Xenopus. Dev. Biol. 177, 96–103. 6. Nieuwkoop, P. D. and Faber, J. (1967) Normal Table of Xenopus laevis (Daudin). NorthHolland, Amsterdam. 7. Tonissen, K. F., Drysdale, T. A., Lints, T. J., et al. (1994) XNkx-2.5, a Xenopus gene related to Nkx-2.5 and tinman: evidence for a conserved role in cardiac development. Dev. Biol. 162, 325–328. 8. Klymkowsky, M. W. and Hanken, J. (1991) Whole-mount staining of Xenopus and other vertebrates. Methods Cell Biol. 36, 419–441. 9. Harland, R. M. (1991) In situ hybridization: an improved whole-mount method for Xenopus embryos. Methods Cell Biol. 36, 685–695.

Craniofacial Development

45

6 Craniofacial Development and Patterning Harold Slavkin, Glen Nuckolls, and Lillian Shum 1. Introduction During early craniofacial development, cranial neural crest cells emigrate from segmentally distinct divisions of the hindbrain (rhombomeres) to populate the various branchial arches and subsequently differentiate into multiple neuronal and non-neuronal cell lineages of the head and neck region. Organ culture of the mandibular portion of the developing first branchial arch (1,2) facilitates investigations of molecular signaling events of crest cell differentiation during chondrogenesis, osteogenesis, odontogenesis, and myogenesis (3–5). In addition, organs explanted in culture are assessable physically by microdissections, implantations, transplantations, and microinjections, and can be readily manipulated by antisense oligonucleotides strategies, viral delivery, and immunoperturbation methodologies. Cell growth and differentiation are often examined at the animal model level using transgenic overexpression and targeted disruption approaches or at cell culture level using transfection and infection techniques. However, organ culture methods provide unique opportunities for the examination of time- and position-dependent differential gene expression related to pattern formations and instructive tissue interactions. Explant organ culture circumvents problems with whole animal modeling, such as lengthy and labor-intensive preparations and early embryonic lethality. It also avoids many of the cell culture artifacts, such as adherence to plasticware and loss of normal cell–cell and/ or cell–extracellular matrix interactions. This protocol describes a methodology that employs a serumless, chemically defined medium permissive for normal differentiation programs; both morphogenesis and cytodifferentiation are comparable to in vivo development. With the absence of confounding variations introduced by serum factors, this protocol is particularly attractive for studies of endogenous growth factors and their cognate receptors. Organ culture coupled with implantation of growth-factor-soaked beads allow for focal delivery of exogenous ligands, which offers a number of approaches. Therefore, this protocol enables a venue for developmental biologists interested in studies of intrinsic morphoregulatory programs during critical stages of early vertebrate embryogenesis.

From: Methods in Molecular Biology, Vol. 136: Developmental Biology Protocols, Vol. II Edited by: R. S. Tuan and C. W. Lo © Humana Press Inc., Totowa, NJ

45

46

Slavkin, Nuckolls, and Shum

2. Materials 2.1. Organ Culture 1. Mouse embryos: Timed pregnant female mice can be purchased from laboratory animal suppliers or bred at in-house facilities. Males and females are caged together overnight; the presence of the vaginal sperm plug on the next morning indicates mating and the day designated as gestation day 0 (see Notes 1–3). 2. Microdissecting instruments: One pair of surgical scissors, one pair of blunt forceps, one pair of spring scissors, two pairs of fine forceps #55, and disposable microscalpels. 3. BGJb medium: Fitton-Jackson modification (cat. no. 12591-038; Life Technologies, Gaithersburg, MD). Basic BGJb medium can be purchased, but the culture medium should be made just prior to culture. Aliquot basic BGJb medium in 10-mL aliquots and store at 4°C. Just prior to use, supplement with 0.1 µg/mL ascorbic acid, 100 U penicillin, and 100 U streptomycin. Work in the laminar flow hood to maintain sterility. 4. Hanks’ balanced salt solution (HBSS): Without phenol red (cat. no. 14025-092; Life Technologies). 5. Supporting filter: Type AA, 0.8-µm pore size (cat. no. AABP-047-00; Millipore, Bedford, MA). Filters are supplied as 47-mm diameter discs. First, stack approx 5 filters together, use a hole punch to punch out 6-mm diameter discs and collect them in a 500-mL glass beaker. Wet and rinse the filters with 5 changes of 200-mL distilled, deionized water (ddH2O). Boil the filters in 200 mL of ddH2O for 5 min and rinse in another two changes of 200 mL ddH2O. Rinse once with 70% ethanol. Work in a laminar flow hood to ensure sterility. Pour the filters and ethanol in several 10-cm diameter petri dishes. Aspirate ethanol from the dishes. Dry the filters by removing the dish cover and allowing the remaining ethanol to evaporate. These filters can be stored at room temperature (RT) indefinitely if maintained sterile. 6. Supporting stainless steel grid: 30-mesh, openings 0.015 in. (cat. no. 5335-00-B55-0105; Cambridge Wirecloth Company, Cambridge, MD). Steel mesh is supplied as sheets. Use metal-cutting scissors to cut square grids of 2-cm sides. Cut a small notch on one side. Clean the grids thoroughly with water and detergent after each use and store in 70% ethanol. 7. Gauze: 2 × 2 in., 4-ply (cat. no. 7632; Johnson and Johnson, Arlington, TX). 8. Organ culture dish: 60 × 15-mm style with center well (cat. no. Falcon 3037; Becton Dickinson, Franklin Lakes, NJ). 9. Tissue culture dish, 10-cm diameter: 100 × 20-mm style (cat. no. Falcon 3003; Becton Dickinson). 10. Dissection dish: Tissue culture dish coated with a 5-mm thick 2% agar on the bottom of the dish. Melt 2 g of agar in 100 mL ddH2O by gentle boiling in a 500-mL glass beaker; avoid foaming and overflowing. Swirl to melt the agar completely, and let sit at RT to cool. In the laminar flow hood, pour approx 10 mL of melted agar into each dish to coat the dish bottom, and allow to cool completely. Seal the dish with Parafilm, store at 4°C, and use within 1 mo. 11. Sterile transfer pipets: (cat. no. 202-1S; Samco Scientific, San Fernando, CA). 12. Water-jacketed carbon dioxide incubator set to 37°C and 5% carbon dioxide.

2.2. Bead Implantation 1. Beads: Affi-Gel Blue Gel at 100–200 mesh (cat. no. 153-7302; BioRad, Hercules, CA). Store at 4°C. 2. Siliconized microfuge tubes: 1.5 mL (Fisher Scientific Co., Pittsburgh, PA). 3. Bovine serum albumin (BSA): Fraction V (cat. no. A1933; Sigma, St. Louis, MO), cellculture tested. Store powder at 2–8°C.

Craniofacial Development

47

4. BSA-coated pipet tips: Prepare 1 mg/mL BSA solution in distilled, deionized water. Coat the inner barrel of pipet tips with BSA by pipetting the BSA solution up and down 2–3 times. Expel the solution completely from the pipet tips and let dry at RT in a laminar flow hood. 5. Phosphate buffered saline (PBS) at pH 7.4. 6. Manipulation of glass capillary for micropipet: Hard glass capillary, 75 mm in length (Drummond Scientific; Broomall, PA). Using a small flame from a bunsen burner, rotate the glass capillary and heat to soften the middle portion. Remove from flame and pull the ends apart, thus drawing the capillary out to a finer diameter to be used as micropipet. Break off the tapered end gently and smoothen the jagged edges by gently flaming the ends. Monitor the inner diameter under a stereomicroscope. The inner diameter of the capillary should be approx 50% of the diameter of the beads used for optimum holding and ease of manipulation. Sterilize by standard autoclaving. 7. Micropipet: Mouth-controlled micropipet can be assembled by connecting the mouthpiece (cat. no. 258616; Curtin Matheson Scientific, Jessup, MD), tubing, glass Pasteur pipet, micropipet holder, and micropipet.

3. Methods

3.1. Organ Culture 1. Assembly of the organ culture setup (Fig. 1A,B). This should be performed in the laminar flow hood and completed prior to the beginning of dissection. Place one piece of gauze and 6 mL of ddH2O in a 10-cm diameter tissue culture plate so that the piece of gauze is thoroughly soaked. Place one organ culture dish on top of the piece of gauze and 3 mL of ddH2O in the outer compartment of the organ culture dish. Use forceps to pick up one supporting grid, drain off excess ethanol, flame to dry and sterilize, and allow to cool briefly. Place this supporting grid over the inner compartment of the organ culture dish. Using a sterile transfer pipet, place enough culture medium, approx 1.5 mL, into the inner compartment so that the meniscus of the medium is even with the top of the grid. Replace the covers and allow the setup to equilibrate to 37°C and 5% CO2 in the incubator. 2. Sacrifice gestation day 10 timed-pregnant female mice by cervical dislocation or carbon dioxide inhalation. Wet the abdominal area with 70% ethanol for sterilization and open the abdomen with bilateral incisions through the skin and body wall. Displace the bowels to reveal the uterus. Excise the bicornate uterus and place immediately in ice-cold HBSS. Rinse the uterus several times with ice cold HBSS to remove as much blood as possible to create a clearer field of microdissection. 3. Isolate each embryo from the implantation site by opening the site with forceps and spring scissors, extracting the embryo, and cleaning away from extraembryonic membranes. Stage the embryos according to Theiler stages (1972). The optimum stage for mandibular process explant is stage 18, with 40–44 somite pairs (Fig. 1C,D) (see Notes 4–6). 4. Microdissect the mandibular portion of the first branchial arch in the dissection dish with HBSS by making the following four incisions: a. Transverse cut immediately superior to the heart to separate the cranial region from the trunk region (Fig. 1E). b. Transverse cut between the first and second branchial arch (Fig. 1F). c. Transverse cut between the maxillary and mandibular portion of the first branchial arch (Fig. 1G). d. Frontal cut to separate the mandibular process from the neural tube (Fig. 1H) (see Notes 7–9).

48

Slavkin, Nuckolls, and Shum

Fig. 1. Microdissection and organ culture assembly. Schematic representation (A), and actual image (B) of the organ culture system illustrate the explant being supported by filter placed on grid and cultured at the atmosphere-medium interface in the culture dish. Water in the outer compartment of the culture dish and water-soaked gauze placed in the outer tissue culture dish serve to maintain optimum humidity for growth and development of the explant. The explant is oriented with the oral epithelium facing upward (B, inset). The optimum stage for collecting mandibular process explants from mouse embryos is Theiler stage 18; 40–44 somite pairs shown at the (C) lateral and (D) frontal views. The mandibular process (D, boxed) is isolated by a series of four incisions (E–H, see text) using fine forceps and

Craniofacial Development

49

5. Select a piece of filter, wet and submerge it into HBSS in the dissection dish. Gently place the mandibular process onto the filter with the oral epithelial side facing up. Use two pairs of fine forceps to pick up the filter with the mandible and raise out of the HBSS, taking care to maintain the filter at a horizontal plane and to not dislodge the orientation of the mandibular process. Place the mandibular process explant on the supporting grid of the organ culture assembly (Fig. 1B, inset) (see Notes 10–15). 6. The assembly can be returned to the incubator. Change to fresh medium every 2 d. Aspirate and replace medium by using a Pasteur pipet accessing the inner compartment of the organ culture dish through the notched side of the supporting grid. Adjust the amount of medium so that the mandibular explants are placed at the atmosphere-medium interface to achieve optimum growth and development (see Note 16).

3.2. Bead Implantation (See Notes 17–19) 1. Preparation of beads: Growth-factor-soaked beads are prepared fresh on the day of use. Swirl the bottle of Affi-Gel Blue beads to resuspend. Take 10 µL of bead suspension and place it in a siliconized microfuge tube. Add 100 µL of PBS into the tube, mix, and allow to sit at RT for 5 min. Collect the beads by centrifugation at 14K for 5 s. Aspirate the supernatant. Repeat washing with PBS once more. Work in the laminar flow hood to maintain sterility from here on. Using a BSA-coated pipet tip, add 5 µL of a 100 ng/µL solution of BMP4 to the beads, mix, and allow to incubate at RT for at least 1 h. 2. With fine forceps, make a small well in the agar of the dissection dish. Cover the agar with a thin layer of BGJb medium, approx 1 mL. Transfer the beads into the well. 3. From the incubator, retrieve one specimen of mandibular process with the filter and place it in the dissection dish. 4. Under the dissecting microscope, using gentle suction on the mouth-controlled micropipet, pick up and hold one bead. 5. Position the tip of the micropipet with the attached bead to the desired point of implantation. Release suction and gently push the bead through the epithelium with the micropipet. Return the specimen to the incubator.

4. Notes 1. Handling of the animals and disposition of carcasses and tissues should be performed in accordance with Animal and Care Guidelines and the institution’s Animal Study Protocols. 2. If timed pregnant mice are purchased from a vendor, then the investigator should ascertain the protocol of counting gestation days as determined by the vendor. Most vendors designate the day of finding a vaginal sperm plug in the female as day 0 of gestation. However, day 0.5 and day 1 are designated by others. The method of assignment is neither universal nor standardized. 3. Frequently, even when particular gestation stage embryos are requested, there is a wide range of variation in actual stage of the embryos among litters and within a litter. The variation can be as much as 12 h of development. Some investigators breed animals for a limited time in their laboratory vivarium (e.g., caged together for only 2–3 h). However, the time of fertilization is often dependent on the time of ovulation (which, in turn, is

microscalpels. 1, frontal process; 2, medial nasal process; 3, lateral nasal process; 4, maxillary process; 5, mandibular process; 6, second branchial (hyoid) arch; 7, heart; 8, forelimb bud; 9, hindlimb bud.

50

4.

5.

6.

7.

8.

9.

10.

11.

12.

Slavkin, Nuckolls, and Shum dependent on the light–dark cycle of the housing facility) and not the time of mating. Therefore, the development variation is an unavoidable and inherent factor in the experiment. The investigator is encouraged to determine the exact stage of the embryos by external morphology, such as employing the number of somite pairs according to the Theiler staging system (7). In mandible explant cultures, the ideal initial stage is between 40 and 44 somite pairs, corresponding to Theiler stage 18. The mandibular process develops and produces multiple phenotypes, including cartilage, osteoid material, incisor and molar tooth buds, and muscular tongue, all within 9 d in serumless medium (Fig. 2A–H). These phenotypes are detected by in 3–6 d of culture. After 9 d, the explant can be kept alive but without further differentiation. For example, the tooth organs will reach the bell stage (but not beyond that) regardless of the length of culture. Therefore, most experiments are terminated after a maximum of 9 d. In addition to the mandibular processes, many other organs from the craniofacial region have been cultured successfully in the system; rhombomeres (8), inner ear (9), palatal shelves (10), and tooth explants (11). Organs from the trunk region are also used; limb bud (12), lung (13), and metatarsal explants (14). Several variations in methodology include the stage of embryo the explant is isolated from, the method of microdissection, and duration of culture. Rhombomeres are the early segmentations of the hindbrain, isolated from E8 embryos. Because of the early stage of these tissues, explant cultures require serum-containing medium for optimal growth and differentiation. The explant culture system not only supports normal growth and development but also allows for the study of dysmorphology when coupled with perturbation strategies such as the use of antisense oligonucleotides (Fig. 3). Microdissection is usually done under a stereomicroscope in a laminar flow hood. However, if the working surface is sufficiently sterilized by thorough wiping with 70% ethanol, the procedure can be performed on a regular laboratory bench. Instruments are thoroughly sterilized by dipping in 70% ethanol and then flaming to dryness and allowed to cool. Often, the flaming may cause discoloration of the instrument without adverse effects to the microdissection. Instruments can also be sterilized by standard autoclaving. Inherently, there is a small but noticeable variation in the amount of tissue isolated for cultures; that is, where exactly to make the cut. Therefore, we suggest that for one complete set of experiments, either only one person should perform the microdissection, for consistency and reproducibility, or efforts should be made to calibrate several scientists in microdissection or both. The mandibular process should be explanted into culture with the oral epithelial side up. This orientation may be challenging to identify for the beginner. One clue is to use the cut surfaces between the maxillary and mandibular process as a guide, because these should be facing up as well, aligned with the oral side. During the step when a cut is made to separate the first and second branchial arches, we suggest that the dissection include a small amount of tissue from the second arch with the explant (Fig. 2A). Empirical data revealed that the inclusion of second arch tissue enhances the symmetrical development of the mandible and Meckel’s cartilage. In one culture assembly, six filters are usually placed on the supporting grid. With some careful arrangement, up to eight filters can be accommodated. On each filter, a maximum of three mandible explants can be supported. If one were to design an experiment that would require a large number of explants, then arranging more mandible explant

Craniofacial Development

51

Fig. 2. Phenotypic differentiation of the mandibular process cultured in serumless, chemically defined medium is comparable to in vivo development. The isolated mandibular process explant at the beginning of culture (A) consists of a simple cuboidal epithelium enclosing a population of ectomesenchymal cells derived from cranial neural crest (D). After 3 d of culture (B), a rudiment of Meckel’s cartilage (mc) at the posterior lateral aspect of the explant is readily observable at the macroscopic level. This unique cartilage further grows into symmetrically positioned rod-shaped structures by 7 d of culture (C). The tongue (tg) is centrally placed. Histological analyses (E–H) illustrate additional structural differentiation, such as the developing tooth bud, which is composed of the dental epithelium (de), surrounding condensing dental mesenchyme (dm), and osteoid materials (os) subsequently to be the mandibular bone deposited adjacent to Meckel’s cartilage. Myoblasts and myotubes can be seen in the tongue. oc, oral cavity; f, filter; ms, median sulcus. (I) Exogenous BMP4 delivered by Affi-Gel Blue beads induces ectopic cartilage formation in vitro. BMP4-soaked bead (b) is implanted in the mandibular process at the beginning of culture. (J) After 6 d of culture, ectopic cartilage surrounding the bead (b) can be detected by whole-mount alcian blue staining. (K) Toluidine blue staining of histological sections demonstrate for cartilage-specific glycosaminoglycans surrounding BMP4-soaked beads (b). per culture assembly would be more convenient during subsequent handling. However, if more than 10 mandibles are placed in one assembly, then daily medium change would be required.

52

Slavkin, Nuckolls, and Shum

Fig. 3. Study of mandibular process dysmorphology using explant culture system coupled with antisense deoxynucleotide strategies (refs. 4,5). Mouse E10-stage mandibular processes were explanted into culture in the presence of 30 µM antisense oligonucleotides directed against epidermal growth factor (EGF), (B), transforming growth factor-beta 1 (TGFβ1) (D), TGFβ2 (F), or TGFβ3 (H). After 9 d in culture, growth and development of Meckel’s cartilage or tooth buds were compared with the respective sense oligonucleotide-treated controls (A, C, E, G). Antisense inhibition of EGF induced fusilli-form dysmorphogenesis of Meckel’s cartilage (B)

Craniofacial Development

53

13. HBSS can be purchased in a variety of modified components. We found that HBSS without phenol red was particularly suitable for use as a dissection buffer, because the solution is clear and allows for better visibility. 14. Dissection is performed using a dissection dish coated with a thin layer of agar to preserve the sharp cutting edge of the microscalpel during the duration of each microdissection session. It also provides a clean cut through of the tissues compared with dissecting against a hard plastic surface of the plate. To provide maximum visual contrast for the identification of embryonic parts during microdissection, it is possible to add 0.5 mL of black India ink into the agar during preparation of the dissection dish. 15. During the preparation of supporting filter, always wear gloves, as grease from the scientist’s fingers causes contamination and uneven wetting of the surface. The particular type of filter is selected because the dark color surface provides visual contrast to the otherwise opaque explants. 16. After several days of culture, the mandible explants will have enough fibroblastic outgrowth so that the explants become adherent to the filters. It is not necessary to separate the explant from the filter for fixation for subsequent histological evaluation, because the filter is porous and that the tissue is thin. In fact, the filter provides a useful substrate for handling the mandibles and for orientation to define the plane of section for histology. 17. The mesh size of Affi-Gel Blue beads corresponds to effective diameter of beads of 50–75 µm. The gel is stored in PBS buffer with sodium azide as preservative. Therefore, do not breathe vapor from the solution. The gel has a shelf life of 1 yr when stored at 2–8°C. The matrix is cross-linked agarose coupled to Cibacron blue dye F3GA. The dye provides active binding sites to proteins and peptides with a capacity of more than 11 mg/mL. AffiGel Blue beads have a high affinity for proteins and therefore are suitable for soaking and delivering many other different peptide growth factors in addition to BMP4 used in this protocol’s investigations (Fig. 2I–K). 18. Growth factors are small polypeptide molecules that have a high affinity to plastic. Therefore, to minimize the loss of exogenous proteins through “stickiness” to handling plasticware, we use siliconized microfuge tubes and BSA-coated pipet tips. Wherever possible, also add carrier BSA to a concentration of 0.1 mg/mL to growth-factor solutions and store growth factors in small aliquots at a high stock concentration. Dilute to working concentration just before use. Avoid freeze–thaw cycles. 19. During microdissection and bead implantation, it is essential to perform the procedures as swiftly as possible. Dissection should be done using ice cold buffer. Viability of the explants decreases with exposure to suboptimal temperature, atmosphere, and medium.

Acknowledgment We thank Pablo Bringas, Jr. for his participation in the development of the organ culture protocol and Kazuaki Nonaka and Ichiro Semba for their contributions to the development of the bead implantation procedures. Human recombinant BMP4 was supplied by Genetics Institutes. This work is supported by NIH funding Z01-AR41114.

and reduction in tooth bud size (data not shown). Abrogation of TGFβ1 resulted in malformation of the rostral region of Meckel’s cartilage and caused the appearance of inward curvature (ic) between the rostral and mid-segment (D). TGFβ2 antisense treatment produced a threefold enlargement of tooth buds (tb) and accelerated the development to cap stage (F). TGFβ3 abrogation resulted in reduction in length and thickness of Meckel’s cartilage (H, arrow).

54

Slavkin, Nuckolls, and Shum

References 1. Slavkin, H. C., Bringas, P., Jr., Sasano, Y., and Mayo, M. (1989) Early embryonic mouse mandibular morphogenesis and cytodifferentiation in serumless, chemically-defined medium: a model for studies of autocrine and/or paracrine regulatory factors. J. Craniofac. Genet. Dev. Biol. 9, 185–205. 2. Slavkin, H. C., Sasano, Y., Kikunaga, S., Bessem, C., Bringas, P., Jr., Mayo, M., Luo, W., Mak, G., Rall, L., and Snead, M. L. (1990) Cartilage, bone and tooth induction during early embryonic mouse mandibular morphogenesis using serumless, chemically-defined medium. Connect. Tissue Res. 24, 41–51. 3. Mayo, M. L., Bringas, P., Jr., Santos, V., Shum, L., and Slavkin, H. C. (1992) Desmin expression during mouse tongue morphogenesis. Int. J. Dev. Biol. 36, 255–263. 4. Shum, L., Sakakura, Y., Bringas, P., Jr., Luo, W., Snead, M. L., Mayo, M., Crohin, C., Millar, S., Werb, Z., Buckley, S., Hall, F. L., Warburton, D., and Slavkin, H. C. (1993) EGF abrogation induced fusilli-form dysmorphogenesis of Meckel’s cartilage during embryonic mouse mandibular morphogenesis in vitro. Development 118, 903–917. 5. Chai, Y., Mah, A., Crohin, C., Groff, S., Bringas, P., Jr., Le, T., Santos, V., and Slavkin, H. C. (1994) Specific transforming growth factor beta subtypes regulate embryonic mouse Meckel’s cartilage and tooth development. Dev. Biol. 162, 85–103. 6. Tabata, M. J., Kim, K., Liu, J. G., Yamashita, K., Matsumura, T., Kato, J., Iwamoto, M., Wakisaka, S., Matsumoto, K., Nakamura, T., Kumegawa, M., and Kurisu, K. (1996) Hepatocyte growth factor is involved in the morphogenesis of tooth germ in murine molars. Development 122, 1243–1251. 7. Theiler, K. (1989) The House Mouse: Atlas of Embryonic Development. Springer-Verlag, New York. 8. Chai, Y., Bringas, P., Jr., Shuler, C., Devaney, E., Grosschedl, R., and Slavkin, H. C. (1997) An early embryonic mouse mandibular culture model permits the study of cranial neural crest cell migration and transcription factor in regulating mouse tooth development during craniofacial morphogenesis. Int. J. Dev. Biol. 42, 87–94. 9. Hoffman, D. S., Bringas, P., Jr., and Slavkin, H. C. (1996) Co-culture of contiguous developmental fields in a serumless, chemically-defined medium: an in vitro model permissive for coordinate development of the mouse ear. Int. J. Dev. Biol. 40, 953–964. 10. Shuler, C. F., Guo, Y., Majumder, A., and Luo, R. Y. (1991) Molecular and morphologic changes during the epithelial-mesenchymal transformation of palatal shelf medial edge epithelium in vitro. Int. J. Dev. Biol. 35, 463–472. 11. Evans, J., Bringas, P., Jr., Nakamura, M., Nakamura, E., Santos, V., and Slavkin, H. C. (1988) Metabolic expression of intrinsic developmental programs for dentine and enamel biomineralization in serumless, chemically-defined, organotypic culture. Calcif. Tissue Int. 42, 220–230. 12. Canoun, C., Ma, C., Halpern, D., Shum, L., Bringas, P., Jr., Sank, A., and Slavkin, H. C. (1993) Endogenous epidermal growth factor regulates limb development. J. Surg. Res. 54, 638–647. 13. Jaskoll, T. F., Johnson, R., Don, G., and Slavkin, H. C. (1986) Embryonic mouse lung morphogenesis in serumless, chemically-defined medium in vitro. Prog. Dev. Biol., part A, 381–384. 14. Minkin, C., St. James, S., Tao, H., Yu, X., Pockwinse, S., MacKay, C., and Marks, S. C., Jr. (1991) Skeletal development and formation of osteoclast-like cells from in situ progenitors in fetal mouse metatarsals cultured in chemically defined medium. Bone Mineral 12, 141–155.

Skeletal Morphogenesis

55

7 Craniofacial Skeletal Morphogenesis In Vitro Roy C. Ogle 1. Introduction All the bones of the craniofacial skeleton are united by the fibrous sutures, and those of the neurocranium are lined additionally by the dura mater. The sutures and dura mater are not only integral structural elements of the formed skeleton, but during morphogenesis they also are sites of appositional growth of the membranous bones of the vault and inductive interactions that regulate the process of suture obliteration, respectively (1–5). When sutures fail to form or are prematurely fused prior to cessation of rapid brain growth, morphogenesis of the entire head is severely altered (6). Although a variety of approaches employing transplantation of embryonic rudiments and surgical perturbation of elements of the developing craniofacial skeleton have demonstrated that cellular interactions among the perisutural tissues are critical to morphogenesis (1), recent development of media and methods for growing fetal rat and mouse calvaria in vitro from non-mineralized rudiments may allow identification of the molecules and mechanisms underlying normal craniofacial development (3,7–11). The cranial vault sutures of rats and mice form during a 72-h period immediately prior to birth. At fetal day 18 (F18) in rats and a day earlier in mice, calcification of the bones is restricted to a small ossification center in each bone separated widely by embryonic mesenchymal tissues. Between each of the paired frontal and parietal bones the coronal sutures develop, with overlapping bones and fibrous sutures that remain open throughout the lives of rodents. Between the frontal bones the intrafrontal suture is formed, which is a butt or nonoverlapping suture that fuses in the third week of life. As described in this protocol, culture of the calvaria under defined conditions, allowing both mineralization and formation of tissues as seen in vivo, can be used to identify the factors originating in the dura mater and developing bones that serve as signals guiding the component processes of morphogenesis—proliferation, differentiation, and apoptosis (see Fig. 1). This method has proven amenable to approaches with antisense oligonucleotide, antibody and pharmacological inhibition of growth factors and their receptors to perturb elements of the system, as well as attempts to reconstitute the system by supplementing defined media with purified factors. An example of the use of a neutralizing antibody to one of the fibroblast growth factor receptors

From: Methods in Molecular Biology, Vol. 136: Developmental Biology Protocols, Vol. II Edited by: R. S. Tuan and C. W. Lo © Humana Press Inc., Totowa, NJ

55

56

Ogle

Fig. 1. Development of the coronal suture region of the rat calvaria in vitro. Photomicrographs (Hoffman modulation contrast optics) of frozen sections (7 µM) stained with Alizarin red to visualize mineralized bone. (A) By 24 h in culture, the frontal and parietal bones have begun to overlap and are partially calcified. (B) At 120 h, the bones and suture have continued growing. Bone mineralization and thickness have increased and the suture remains nonossified. If the dura mater is removed prior to culture, the bones fuse (not shown, similar to Fig. 2B) unless appropriate factors are supplied in the defined medium. Bars = 100 µM.

(FGFR) is included to illustrate this approach. This system may also be employed to investigate morphogenesis of the axial and appendicular skeletons and other events of appropriate or abnormal bony fusion, such as the formation of the digits. 2. Materials 1. Serum-free culture medium (see Notes 1 and 2): Dulbecco’s high-glucose minimum essential medium with pyruvate (DMEM; Gibco, BRL, Gaithersburg, MD) containing 1 µg/mL gentamicin, 5 mM glutamine, 1X nonessential amino acids (100X stock; Gibco-BRL), 1 mM insulin, transferrin, and selenium (ITS+; Collaborative Research, Inc., New Bedford, MA), and 3 mM inorganic phosphate (stock solution = 0.1M NaH2PO4, pH 7.0). Cultures were supplemented daily with 100 µg/mL ascorbic acid (stock solution = 0.1M sodium ascorbate in H2O; Sigma Chemical Co., St. Louis, MO). 2. F19 Sprague-Dawley rat fetuses, 1-day-old (N1) or 17-day-old (N17) male rat pups from laboratory breeding colony. 3. Hank’s balanced salt solution (HBSS): 0.14 g CaCl2, 0.4 g KCl, 0.06 g KH2PO4. 0.1 g MgCL2(6H2O), 0.1 g MgSO4 (7H2O), 8 g NaCl, 0.35 g NaHCO3, 0.06 g Na2HPO4 (2H2O). and 1.25 g glucose in 1 L dH2O. 4. Cultureware: 24-well culture dishes and 0.45-µm polycarbonate membrane inserts (Corning-Costar, Corning, NY). 5. Carnoy’s fixative: 60 mL absolute ethanol, 30 mL chloroform, and 10 mL acetic acid. 6. 0.2% Alizarin Red-S (Sigma) 0.2 g in 100 mL dH2O plus 1 drop concentrated ammonium hydroxide (Fisher Scientific Co., Pittsburgh, PA). 7. Decalcification solution: ready-to-use EDTA/HCl solution (Stephens Scientific, Riverdale, NJ).

Skeletal Morphogenesis

57

8. Statistical analysis software: SAS (SAS Institute, Cary, NC). 9. Calcium analysis: 5% trichloroacetic acid (Fisher Scientific), Sigma kit #587-A and Sigma calcium standard #360-5.

3. Methods 1. Preparation of explants for culture (see Note 3): F19 Sprague-Dawley rat fetuses or (N17) male neonates are used. Male and female rats are caged as mating pairs in the late afternoon and females are checked for plugs the following morning. On discovery of a plug, males and females are immediately separated. Day of plug is denoted as day 0. At the appropriate developmental stage, pregnant females are killed by overadministration of Halothane and pups aseptically removed and placed on ice prior to removal of calvaria. The scalp is opened and whole calvaria removed from pups and placed in HBSS on ice. The frontal and parietal bones, including presumptive coronal suture, sagittal, and intrafrontal sutures are carefully dissected away from the surrounding tissues. Using a dissecting microscope, the dura mater is removed from those calvaria to be cultured denuded of dura mater, with care being taken to eliminate all traces of dura mater without damaging the underlying suture. In experiments focusing only on the coronal suture, F19 calvaria are divided in half by cutting through the sagittal and intrafrontal sutures. In cases in which the intrafrontal is of sole interest, the N17 calvaria are cut through the coronal suture. 2. Culture and harvesting of calvaria: Although calvaria were typically cultured with dura mater intact, in reconstitution experiments, calvaria with dura mater removed are used, occasionally separated from (but in liquid contact with) the dissected dura mater by a 0.45-µm polycarbonate membrane insert. Calvaria are placed dura side down into 24-well culture dishes and covered with 400 µL medium (see Note 4). If antibodies or other test substances are to be added, they and vehicle controls, if needed, are added at this point. In the example shown in Fig. 2B, 4 µg per 400 µL of medium of neutralizing IgG against FGFR2 (Santa Cruz Labs) was employed. All groups receive serum-free medium, which is replaced by one half with fresh medium and supplemented daily with 100 µg/mL ascorbic acid. Individuals from each group are harvested at 24-h or more frequent intervals and prepared either for histology or calcium analysis. 3. Preparation of tissues for histology: Explants are fixed in Carnoy’s fixative overnight and transferred to 70% ethanol, stained with 0.02% Alizarin red overnight (see Note 5), embedded in OCT compound (Tissue-tek), and frozen sectioned (7 µM). In situations requiring optimal morphology, specimens are decalcified (20 min) with EDTA/HCl, and processed for paraffin sectioning. Typically, thirty-six 5-µM sagittal sections are cut through the central portion of each suture and stained with hematoxylin and eosin (H & E). Assessment of the patent state of the suture serves as a bioassay for establishing the ability of various conditioned media, their fractions, and known growth factors to promote fusion in the coronal or prevent sutural obliteration of the intrafrontals. 4. Histomorphometry and statistical analysis (see Note 6): Sections are randomized according to random numbers generated by SAS (SAS Institute, Cary, NC) and examined by two independent observers such that observers are blinded to experimental manipulation. Sections exhibiting poor histology or incorrect plane of section are excluded from analysis. Each section is scored for degree of osseous obliteration of sutures (intact sutures = 3; bone fronts touching, but not fused = 2; bone fronts fused in region of suture = 1) and thickness of bones. Because osseous obliteration of sutures occurs in several ways, intact sutures are scored as 100% and degree of osseous obliteration calculated as a percentage thereof. Results are tested by analysis of variance as previously described (1,3). Data are presented as mean + SEM.

58

Ogle

Fig. 2. Development of rat calvaria in vitro—effects of neutralizing antibody to FGFR. Photomicrographs (Hoffman modulation contrast optics) of frozen sections (7 µM) stained with Alizarin red to visualize mineralized bone. (A) Control calvaria cultured for 48 h as in Fig. 1. (B) Calvaria cultured in presence of purified rabbit immunoglobulin G against FGFR 2 (4 µg in 400 µL) for 48 h. Note the fusion of the parietal and frontal bones in the region indicated by arrowheads and the apparent increased growth of bone and fibrous tissues. 5. Calcium analysis: Calcium is extracted from calvaria with two 1-h incubations in 5% trichloroacetic acid (1 mL each), which are pooled for analysis using Sigma kit #587-A and Sigma calcium standard #360-5. The tissues are then washed 3X in acetone and allowed to air dry overnight prior to weighing on a Mettler 6300 balance.

4. Notes 1. A wide variety of methods and additives for culture were tested in developing this methodology. Whereas other media are more commonly used for bone cell culture, the highglucose DMEM used as described worked best on intact calvaria. Inclusion of 1–10% fetal calf serum actually inhibited growth and should be avoided. Use of beta-glycerol phosphate instead of inorganic phosphate and concentrations of phosphate lower than 3 mM in addition to that present in DMEM resulted in ectopic calcification within the periosteum. 2. The source of transferrin in the insulin, transferrin and selenium (ITS+) used in the media is human, and precautions taken when working with human blood products should be observed. This laboratory has used transferrin that was later determined to have been prepared from pooled blood that contained viruses (Creutzfeld-Jacob), although the transferrin preparation was reportedly free of virus. 3. It is essential to take care not to damage dura or suture during dissection. A small amount of dura remaining after removal can be enough to block fusion of the suture, while damage to the suture mesenchyme can prevent its normal development. If litter size is sufficient, several dissected rudiments should be analyzed by histology to assure the quality of dissection. 4. Although 200 µL of medium is sufficient to cover the explants, 400 µL is the minimal amount that gave good reproducible growth. As the explants grow, by 72 h in culture

Skeletal Morphogenesis

59

sufficiently acidic byproducts are secreted into the medium to lower pH as indicated by yellow color of the phenol red indicator dye. In these cases, sterile sodium bicarbonate was added in a dropwise fashion to titrate the medium back to neutrality (cherry red). 5. Alizarin staining of whole skulls did not work well unless the 70% ethanol soak between fixation and staining was used. 6. The type of histomorphometric analysis employed to evaluate suture fusion is very laborious, and others have attempted to evaluate outcomes more rapidly by whole-mount immunostaining or in situ hybridization. That method does not work with the coronal or other overlapping sutures. The geometry of the tissues is such that only cross-sectional analysis indicates the degree of suture fusion.

Acknowledgments The author thanks Lynne Opperman, Ralph Passarelli, Ellen Morgan, Amber Nolen, Chandan Chopra, Anikar Chhabra, Richard Cho, and Mark Reintjes for their roles in developing this protocol. This work was supported by NIH grant DEI0369. References 1. Opperman, L. A., Sweeney, T. M., Redmon, J., Persing, J. A., and Ogle, R. C. (1993) Tissue interactions with underlying dura mater inhibit osseous obliteration of developing cranial sutures. Dev. Dyn. 198, 312–322. 2. Opperman, L. A., Persing, J. C., Sheen, R., and Ogle, R. C. (1994) In the absence of periosteum, transplanted fetal and neonatal rat coronal sutures resist osseous obliteration. J. Craniofac. Surg. 5, 327–332. 3. Opperman, L. A., Passarelli, R. W., Morgan, E. P., Reintjes, M., and Ogle, R. C. (1995) Cranial sutures require tissue interactions with dura mater to resist osseous obliteration in vitro. J. Bone Min. Res. 10, 1978–1987. 4. Opperman, L. A., Passarelli, R. W, Nolen, A. A., Lin, K. Y., Gampper, T. G., and Ogle, R. C. (1996) Dura mater secretes soluble, heparin-binding factors required for cranial suture morphogenesis. In Vitro Cell. Dev. Biol. 32, 627–632. 5. Opperman, L. A., Nolen, A. A., and Ogle, R. C. (1997) TGF-β1, TGF-β2 and TGF-β3 exhibit distinct patterns of expression during cranial suture formation and obliteration in vivo and in vitro. J. Bone Min. Res. 3, 301–310. 6. Enlow, D. H. (1989) Normal and abnormal patterns of craniofacial growth, in Scientific Foundations and Surgical Treatment of Craniosynostosis (Persing, J. A., Edgerton, M. T., and Jane, J. A., eds.), Williams & Wilkins, Baltimore, MD, pp. 83–86. 7. Gronowicz, G., Woodiel, F. N., McCarthy, M.-B., and Raisz, L. G. (1989) In vitro mineralization of fetal rat parietal bones in defined serum-free medium: effect of β-glycerol phosphate. J. Bone Min. Res. 4, 313–324. 8. Chen, T. L. and Bates, R. L. (1993) Recombinant human transforming growth factor β1 modulates bone remodeling in a mineralizing bone organ culture. J. Bone Min. Res. 8, 423–427. 9. Dieudonne, S. C., Foo, P., van Zoelen, E. J., and Burger, E. H. (1991) Inhibiting and stimulating effects of TGF-β1 on osteoclastic bone resorption in fetal mouse bone organ cultures. J. Bone Min. Res. 6, 479–487. 10. Hock, J. M., Canalis, E., and Centrella, M. (1990) Transforming growth factor-β stimulates bone matrix apposition and bone cell replication in cultured fetal rat calvaria. Endocrinology 126, 421–426. 11. Wilkie, A. O. M. (1997) Craniosynostosis: genes and mechanisms. Hum. Mol. Gen. 6, 1647–1656.

Skeletal Morphogenesis

61

8 Skeletal Morphogenesis Stefan Mundlos 1. Introduction The formation of a skeleton with its numerous bones of various shapes and sizes and the growth of these bones from embryonic to adult size is a complex process involving a multitude of genes. The complexity of the process is reflected by the large number of inherited diseases with skeletal phenotypes (in man and mouse) as well as by the everincreasing number of genes shown to be involved in skeletal morphogenesis (for review, see refs. 1,2). The inactivation of genes through transgenic mouse technology has become a popular method of analyzing the function of newly discovered genes. Such mice may have expected or unexpected phenotypes and, considering the large number of genes involved, skeletal alterations are commonly found. Naturally, the question arises, What is the role of this gene in skeletal morphogenesis and how can the phenotype be explained. This chapter will address these questions and will supply the investigator with a conceptual framework for how to investigate skeletal defects in the mouse. Three distinct lineages contribute to the early skeleton. Most cells of the craniofacial bones are of neural crest origin. Neural crest cells migrate from the dorsal side of the neural tube through the sclerotome into the head region. They contribute to the calvarium, midface, mandible and the teeth. Vertebrae and ribs originate from the sclerotome, a structure that forms through the differentiation of somites. The appendicular skeleton is derived from lateral plate mesoderm. Skeletal morphogenesis from these three lineages involves, in principle, four different steps: patterning, organogenesis, growth, and homeostasis (Fig. 1). Patterning determines the future size, shape (gestalt), and number of individual skeletal elements. This is achieved through the expression of patterning genes in progenitor cells long before overt skeletogenesis. Organogenesis involves the generation of bone/cartilage as an organ (skeleton). During this process, undifferentiated cells condense (condensation) according to a preset pattern, differentiate into chondrocytes or osteoblasts (differentiation), and start producing their specific extracellular matrix (histogenesis). These aspects of early skeletogenesis, the “anlage,” represent the outlines of the future skeletal elements. Bones develop either by direct formation of osteoblastic progenitor cells or within a

From: Methods in Molecular Biology, Vol. 136: Developmental Biology Protocols, Vol. II Edited by: R. S. Tuan and C. W. Lo © Humana Press Inc., Totowa, NJ

61

62

Mundlos

Fig. 1. (See color plate 1 appearing after p. 262.) Schematic showing the different steps of skeletal morphogenesis.

temporary framework of hyaline cartilage. The direct formation of bone is called “desmal” or “intramembranous” and produces the flat bones of the skull, part of the clavicle, and the mandible. In contrast, all other bones develop through endochondral bone formation, a process that involves the proliferation and differentiation of chondrocytes to form a temporary cartilaginous template that is subsequently replaced by bone. All longitudinal growth takes place by proliferation and differentiation of chondrocytes. Chondrocytes in the middle of the cartilaginous anlage hypertrophy and their matrix calcifies, while newly differentiated osteoblasts deposit a thin layer of bone around the central part of the anlage. Blood vessels invade this center of primary ossification via the periosteal bone shaft. Some of the invading cells differentiate into hematopoietic stem cells, others into osteoclasts or osteoblasts. In what will become long bones, the primary center of ossification expands toward the two ends. In addition, secondary centers will form at one or both ends. At each side of the primary ossification center, a growth plate forms, highly specialized regions of cartilage where all longitudinal growth takes place via endochondral ossification. Histologically and functionally, growth plate chondrocytes can be subdivided into several zones with specific functions (Fig. 2). Bone mass, shape, and strength are maintained throughout development and adult life through an equilibrium between the forces of bone destruction and those of bone formation. Homeostasis is the process that controls the continuous remodeling of bone. Molecular defects in any of these steps can lead to defects in skeletal morphogenesis. Defects of patterning are characterized by changes in gestalt or number of bones, whereas the remaining skeleton is structurally and functionally normal. Such defects are frequently caused by early patterning genes such as Hox or Pax genes. Defects of growth are characterized by stunted, disproportionate size with the majority of skeletal elements affected. Bone, and/or growth plates show abnormal histology and/or abnormal expression of marker genes. Molecular defects that lead to growth defects are frequently caused by mutations in genes that encode for extracellular matrix (ECM) products or for regulatory signal peptides. Remodeling of bone requires resorption.

Skeletal Morphogenesis

63

Fig. 2. Growth plate with different zones of chondrocyte differentiation. Expression of marker genes is shown on the right side.

Bone resorption is carried out by osteoclasts, highly specialized cell type derived from monocyte precursors. Defects in osteoclast function lead to increased bone density (osteopetrosis). Decreased bones density resulting from increased absorption or decreased production results in osteoporosis. 2. Materials 1. Staining solution: Prepare stock solutions of Alcian Blue (Sigma A3157; Sigma Chemical Co., St. Louis, MO), (0.14% in 60% ethanol) and Alizarin Red (Sigma A5533), (0.06% in 60% ethanol). For 1 L of staining solution, add components in following order and stir: 33 mL stock solution Alcian Blue, 33 mL stock solution Alizarin Red, 230 mL H2O, 130 mL glacial acetic acid, 573 mL ethanol 100%. 2. Fixative: 4% formalin in 1X phosphate buffered saline (PBS), with pH adjusted to 7.2. 3. Van Kossa stain: Prepare solutions of: silver nitrate (Sigma S0139) 1%, pyrogallol (Sigma P0381) 0.5%, and, as fixative, sodiumthiosulfate (Sigma S8503) 5%. Staining is Neutral Red (Sigma N6634). 4. Decalcification: Dissolve EDTA tetrasodium 20% in formalin 4%, and adjust pH to 7.2 with powdered citric acid.

3. Methods In many instances, the first question will be, Is there a skeletal phenotype and how can it be visualized and documented? What the X-ray is in the human, the Alizarin Red stain is in the mouse. The combination of Alizarin Red and Alcian blue allows a differ-

64

Mundlos

ential staining of cartilage (blue) and bone (red). The treatment with KOH and glycerol dissolves muscle and connective tissue but spares bone and cartilage. We use two methods. Method A gives best results for embryos and young mice up to 3 wk of age. Method B stains only bone and is best for adult mice. For best results, mutants should be stained, along with their normal litter mates, at several time points (e.g., E15.5, newborn, 2 wk, adults).

3.1. Preparation and Staining of Skeletons 3.1.1. Method A Embryos of E13.5–E14.5 are isolated and adjacent membranes removed and fixed in ethanol 95%. Older stages are killed and eviscerated. Using a delicate forceps and a dissecting microscope, remove as much of the skin and underlying adipose tissue as possible. The quality of preparations is very dependent on the care taken here. Bone will not stain if covered by skin. In older stages (>1 wk), the skin can be peeled off from tail to head. Carefully remove as much adipose tissue as possible, because it will not be removed by the KOH digestion. Keep specimen in ethanol 95% overnight (3 h minimum). Remove ethanol, keep in staining solution for 2 d. Remove staining solution and clear with KOH. The intensity and duration of KOH digestion has to be adapted to the developmental stage of the embryo. For 3-wk-old mice, we use 2% KOH for 1 d, followed by 1.8% KOH for 3 d, and 0.3% KOH for 4 d. A 1-wk-old mouse will need 2% KOH for 6 h, 1.8% KOH for 2 d, and 0.3% KOH for 1 d. For earlier stages, the times have to be considerably shortened and only 1.8% KOH is used. 3.1.2. Method B This method is used for adult mice only. Skin and eviscerate specimen. Cover specimen with 1% KOH for 3–4 d, followed by 2% KOH for 1 d. Pour out solution, rinse carcass with water, clean tail, cover with 1.8% KOH containing 0.004% Alizarin Red, stain for 1–2 d. Pour out solution, rinse specimen, and clear in (1:1) glycerol: 70% ethanol (several days), transfer into 80% ethanol, and store in 100% glycerol. Once a phenotype has been described and adequately documented, the question How can the phenotype be explained? can be asked. Before starting a set of complicated experiments, it is important to realize what sort of defect the investigator is dealing with. As discussed in the introduction, skeletal defects can be subdivided into defects of patterning, organogenesis, growth, and homeostasis (Fig. 1) Does the mutant show abnormalities in gestalt or number of single bones of embryologically related elements (i.e., vertebrae/ribs, limbs, or cranial skeleton), or are the changes present in the majority of bones regardless of their origin? Is there stunted growth or is the growth potential normal? Is the density of bone normal, increased, or decreased? The answer to these questions leads the way to the experimental design. 3.2. Patterning of the Skeletal Elements If the phenotype turns out to be a defect of patterning, there is little use in looking at histological sections of affected bones. They will be normal. In contrast, whole-mount in situ experiments with genes that are known to pattern this part of the skeleton will be useful to determine where and when alterations occur. The methods used for wholemount in situ hybridization have been described in detail elsewhere (3a). We generally amplify our probes from embryonic RNA with gene-specific primers with a T7 or T3

Skeletal Morphogenesis

65

promoter added to the 3' primer. The PCR product is cleaned on spin columns (Qiagen, Inc., Chatsworth, CA) and used as a template to transcribe Dig-labeled cRNA according to standard protocols (e.g., Genius Kit, Boehringer Mannheim, Mannheim, Germany). Patterning of individual skeletal elements is an extremely complex process regulated by many, mostly unknown genes. It is beyond the scope of this chapter to discuss all the factors, pathways, and mechanisms. It is, however, possible to give a framework that allows the investigator to perform a series of basic experiments that may lead the way to a more detailed analysis. Ribs and vertebrae originate from the sclerotome, a structure derived from the somites. During gastrulation, somites form from the paraxial mesoderm as balls of undifferentiated cells on both sides of the neural tube. Signals from the notochord lead to differentiation of the somite into a dorsal dermomyotome and a ventral sclerotome. Sclerotome cells migrate toward the notochord where they condense to form the vertebrae. The notochord regresses and eventually contributes to the intervertebral discs. Mox1 is a good marker of the early, undifferentiated somite. Mox1 is expressed in all cell types of somites, exhibiting a domain of weak expression in the anterior portion and a domain of strong expression in the posterior portion (3). Sclerotome markers are Pax1, and Pax9 (4), and later in the precondensation and condensation stage, Col2a1. Paraxis can be used as a marker for dermomyotome and sclerotome (5). The myotome can be identified through expression of myogenin and myf5 (6). Notochord tissue can be detected using brachyury (T) (7), or Hnf-3β (8) as markers. Shh is the signal from the notochord that induces the sclerotome and is thus also an excellent marker (9). Pax3 is expressed along the dorsal neural tube and neural crest cells and later, in the lateral half of the dermomyotome (10). The appendicular skeleton originates from a dual contribution of the lateral plate and the somitic mesoderm. After the initial outgrowth of the limb, bud cells from the lateral edges of nearby somites migrate into the limb and contribute to limb muscles, nerves, and vasculature. All other limb tissues, including skeletogenic mesenchyme, derive from lateral plate mesoderm. The skeletal elements of the limbs are laid down sequentially as cartilaginous templates in a proximal-to-distal fashion so that the humeral anlage forms first, followed by the radius and ulna, and finally the digits. Cells of the limb bud are embedded in a three-dimensional structure of signaling molecules that determine their fate (for a detailed description of limb development, see ref. 11). Many of these molecules may be used as markers to detect alterations in expression patterns that may explain the observed phenotype. Fgf8 (early and throughout the ridge) (12) and Fgf4 (later and more posterior) (13) are good markers for the apical ectodermal ridge (AER), a structure of specialized cells at the outer most edge of the limb bud that control outgrowth and proliferation in the proximal-distal dimension. Shh mediates the activity of the polarizing region (ZPA), a group of cells at the posterior margin of the limb bud that establish anteroposterior pattern (13). Hox genes of the A and D clusters show characteristic stage-dependent expression patterns (14) that determine the skeletal pattern of zeugopod (A11, D11) and autopod (A13, D12, D13). The segmentation of the digital anlage into individual phalanges and the separation of digits from one another is achieved through controlled cell death, a process that is likely to be Bmp mediated (15,16).

66

Mundlos

3.3. Organogenesis Signaling centers provide positional information that is recorded on a cell-by-cell basis. With this information, cells are now able to differentiate into specific tissues, a process that results in organogenesis. Genes that are involved in the process of organogenesis are likely to play a role in growth as well. The expected phenotype may therefore involve defects in patterning as well as growth. As a first step in organogenesis, skeletal precursors condense and will then differentiate into chondrocytes or osteoblasts to form the anlage of the future bone. The molecular mechanisms that control condensation and differentiation are largely unknown. Initial adhesion is likely to be mediated by adhesive interactions between cell-surface and extracellular matrix molecules. Transmembrane proteoglycans, such as syndecan-3, have been implicated in this process (17). Recent experiments have shown that the differentiation of skeletal precursor cells into osteoblasts is controlled by Cbfa1, a transcription factor that is expressed in skeletal precursor cells as well as in osteoblasts. Mice with inactive Cbfa1 alleles develop cartilage but no bone. Cbfa1 regulates the expression of many of the bone-related genes such as osteocalcin, collagen type I, and osteopontin (18,19). Whether Sox9, a transcription factor expressed in prechondrocytes and chondrocytes (20), has a similar function in controlling chondrocyte differentiation remains to be determined. Both genes are good markers for skeletal organogenesis. 3.4. Histology of Bones Defects of growth generally affect the growth plate as an organ and result in stunted, disproportionate growth. In this case, histology of growth plates and entire bones at various stages is extremely useful to determine the nature of the defect. Sections of long bones (i.e., humerus or femur) at stages 14.5 and after birth stained with standard H&E will give a first idea whether this is a defect of matrix or a defect of proliferation and/or differentiation. In addition, the material can be used for in situ hybridizations to be discussed. All tissues are fixed in phosphate buffered solution (PBS) 4% formalin. Frozen sections or sections from paraffin-embedded tissue can be used. Paraffin has the advantage that histology is well maintained, paraffin blocks keep forever, and sections can be stored for many months before use (regular stain or ISH). Embryos of less than E17.5 do not need to be decalcified and can be processed as for regular tissue specimens. Older specimens should be decalcified either with 0.2N HCl or in formic acid-sodium citrate until they become soft (depending on the size of the specimen, usually several days). Standard staining procedures include H&E, toluidine blue for staining of cartilage, and van Kossa staining for visualization of calcified tissue (nondecalcified tissue only!). For van Kossa stain, use deparaffinized or frozen sections, stain in 1% silver nitrate for 1–3 min, followed by pyrogallol 30 s. Fix in 5% sodium thiosulfate for 3–5 min, wash in tap water for 10 min, aqua dest. for 30 s, counterstain with neutral red or nuclear fast red. With these sections in hand, the following questions can be asked. What is the overall aspect of the growth plate? Is the cellular architecture maintained, or is there loss of cartilage differentiation? Is the size of the different zones maintained, are zones expanded, or have they disappeared? Is the staining of cartilage or bone matrix altered? The general aspect of histology can be substantiated by in situ hybridization with stage-

Skeletal Morphogenesis

67

specific marker genes. The methods used for in situ hybridization have been described elsewhere in detail (22a). Figure 2 shows a schematic diagram of the different zones of cartilage and bone differentiation as they are generally observed in the growth plate. Chondrocytes of the resting zone are kept in a quiescent, undifferentiated state by expressing fibroblast growth factor receptor 3 (Fgfr3) (21). Growth hormone and IGFs lead to a recruitment of resting cells to the zone of proliferation, where cells rapidly divide. In the transition zone, indian hedgehog and PTHrP regulate, through a feedback loop involving ptc and PTHrP-receptor, the differentiation of proliferating cells into hypertrophic cells (22). Once determined to hypertrophy, chondrocytes start expressing collagen type X. In the lower hypertrophic zone, the extracellular matrix begins to calcify and cells express osteopontin. Collagen type II and aggrecan are expressed throughout the cartilage except in the lower calcified hypertrophic zone. Collagen type I is found in the perichondrium and osteoblasts but not in cartilage. Osteocalcin is a marker for mature osteocytes. Sparc, or osteonectin, is expressed in perichondrium, osteoblasts, and to a lesser extent, in nonhypertrophic cartilage (for a review of the expression of matrix genes, see ref. 23). The testing with these and possibly other genes allows a detailed analysis of growth plate function. If there is evidence for altered extracellular matrix, EM histology of different parts of the growth plate should be performed. Excellent cellular morphology and good fixation of cartilaginous matrix is achieved by ruthenium red fixation, as described by Hunsicker and Herrmann (24). The rate of proliferation can be tested by BrdU incorporation in dividing cells. Inject pregnant females intraperitoneally with 50 µg BrdU (cat, no. 280 879; Boehringer Mannheim) per gram of body weight in PBS. Animals are killed after 2 h, and embryos isolated, fixed, and embedded in paraffin. Sections of growth plates are performed and BrdU-positive cells can be detected by an anti-BrdU-AP monoclonal antibody (cat. no. 1 758 756; Boehringer Mannheim). Labeled cells should be found within the zone of proliferation. Quantification of labeled cells and comparison to controls allows the detection of altered rates of proliferation. Bone homeostasis is dependent on bone production by osteoblasts on one side and bone resorption by osteoclasts on the other. The regulation of this intricate process is largely unknown. Regular histology of entire, decalcified bones will give first evidence of altered bone density and/or structure. In earlier stages, undecalcified sections should be processed for van Kossa stains to document ossification. In osteoporotic bones, the trabecular and, later, the diaphyseal bone will be reduced and thinned. In contrast, in osteopetrosis, much of the marrow will be filled with dense bone, leaving little place for marrow cells. Such alterations can be quantified by counting the number and measuring the width of trabecles in a given area or by measuring diaphyseal bone strength. The number of osteoclasts can be estimated by detection of tartrate-resistant acid phosphatase (TRAP). The most effective and specific method is detection through TRAP-RNA in situ hybridization.

3.5. Primer Sequences The following sequences give 5' and 3' primers to amplify specific probes from Mouse cDNA for in situ hybridization. PCR conditions are: 94°C 4' followed by

68

Mundlos

35X (94°C 1', 60°C 45", 72°C 45"), followed by 72°C 10'. The T7 promoter sequence (5'> CCT ATA GTG AGT CGT ATT AGG 40 C/mmol is common. Tritium must be the isotope used because of its short tracking distance. We use

Analysis of Brain Growth

10. 11. 12. 13.

14.

15.

16.

169

methyl 1', 2',-3H from Amersham (Arlington, Heights, IL), catalog #TRK.565, which is 100–130 Ci/mmol or 37 MBq/mCi. Kodak NTB II emulsion: Catalog # 1654433 from International Biotechnologies, Inc. (New Haven, CT). Siliconizing solution: Prosil 28, organosilane concentrate, product No. 11975-0, Thomas Scientific (Swedesboro, NJ) catalog #7805525. D.P.X. a clear mountant available from Aldrich Chemicals (Milwaukee, WI) catalog #31, 761-6. Embryo instruments: iris scissors; embryo transfer spatula (section lifter) from Carolina Biological Supply Co., Burlington, SC, catalog #K3-62-7510; stainless-steel dissecting probes; watchmaker’s forceps (Dumont #55); medium-point, serrated forceps (high-quality instruments can be purchased from Fine Science Tools, Inc., Foster City, CA), disposable pipets with 7 in, tip; latex bulbs for disposable pipets in 1- and 2-mL sizes. Lab-made instruments: Small glass rods (0.5 mm in diameter) made by pulling out the end of 1-mm glass rods and rounding off the end in a hot flame; hair loops made by attaching a small loop of baby hair to the end of a Pasteur pipet using paraffin. Be sure to use baby hair only because it is extremely fine. Even children’s hair is too coarse. Surgical sponges: Coil up a 1-cm piece of Kimwipe (Kimberly-Clark Corp., Roswell, GA) into a tight wick and hold in place with a hemostat. Microburner for pulling microglass rods and tubes: Attach a 16-gage needle to rubber tubing that is attached to a gas outlet for a standard-sized Bunsen burner. Seal the attachment site with electrical tape. Cautionary note: Be sure that the free end of the needle is pointing into free space, as the flame can be very long (10–12 in.) before adjusting it after you light the needle. Assortment of sterile Petri plates: 90-, 60-, and 30-mm in diameter; standard 3- and 4-in. diameter glass culture dishes. To keep glass dishes sterile, cover the tops with aluminum foil (shiny side out) before autoclaving.

3. Methods

3.1. Embryo Preparation and Morphometrics from Sections 1. Precisely time the age of the embryo by staging it according to the criteria established by Hamburger and Hamilton (8). Counting the number of somites is the most accurate way to stage young embryos. In general, they add a discrete pair every hour between stages 8 and 26 (8). 2. Position the egg on its horizontal axis for at least 15 min to assure rotation of the embryo to the top of the yolk. Crack the shell by hitting the bottom of the egg on a sharp edge (e.g., the edge of a Petri dish). Dump the contents into a culture dish half-filled with physiological saline (p. saline) by placing both thumbs next to the crack and rotating your hands outward from the crack (Note 1). 3. Excise the embryo off the yolk by pinching the vitelline membrane outside of the embryonic disk on the left with a pair of medium forceps while cutting counterclockwise around the outside of the blastodisk, starting on the right side. 4. Transfer the blastodisk to a small Petri dish (60 mm) containing sterile p. saline by using a transfer pipet constructed from a disposable Pasteur pipet. To prepare the pipet, break off the narrow tip and put a 2-mL latex bulb on this end. Use the other end for transferring the embryo. Once transferred, swirl the dish gently to remove excess yolk and the vitelline membrane, which will appear as a translucent film. Then, transfer the embryo to another Petri dish containing p. saline. 5. Flatten the embryo by aspirating off the p. saline. Be sure that the embryo is alive when adding the fixative, Carnoy’s fluid, to the dish. Be careful not to place the fixative directly

170

6.

7.

8.

9.

10.

11.

12.

Desmond and Haas on top of the embryo, as it may cause the embryo to wrinkle. For chick embryos younger than 4 d, be sure that the blastodisk is flat after the fixative is added. Flattening is best achieved by aspirating around the entire circumference of the blastodisk using a disposable pipet. It is important to add ample fixative, dehydrants, and clearing agents to assure adequate processing. This is achieved by adding a volume of fluid that is at least 40X the volume of the embryo (e.g., immerse a stage 10 embryo in 15–20 mL of fluid and a stage 25 embryo in 25–30 mL). Keep the embryos in fixative 10–12 h, being careful to not exceed 24 h. To prevent evaporation of the fixative, place the small Petri dishes inside a larger covered container such as a plastic food-storage dish. Do not cover containers containing fixatives with aluminum foil because the fixatives will cause the foil to disintegrate and the products will fall into the container with your specimens. Remove the Carnoy’s fluid using a disposable pipet. Add 70% ethanol. Swirl the embryo, then aspirate off the ethanol and add fresh 70% ethanol. The embryo may be stored in this solution for months as long as care is taken to avoid desiccation. This is a good time to trim the blastodisk into a rectangular shape. If planning to continue the dehydration series, leave the embryo in this first ethanol solution for 1 h. Ideally, it is best to place the samples on a shaker. You may want to add some eosin stain to the 70% ethanol so that the embryo will be easier to see when processing. The eosin will be removed once the sections are processed through a graded series of ethanols. Continue the dehydration of the embryo by making the following changes of solutions at 1-h intervals: 70% ethanol; two changes of 95% ethanol; two changes of absolute ethanol. Now, transfer the embryo to a glass vial such as a scintillation vial and make the following changes: two changes of histosol/hemode (see Note 2); two changes of 50:50 hemode:paraffin; three changes of pure paraffin (see Note 3). Embed the embryos for sagittal sections, as they are the easiest from which to discern parts of the embryonic brain accurately. Use a dissecting microscope to assure proper orientation of the embryo in the embedding mold (Note 4). Partially fill the mold and place the embryo to the bottom using a fine pair of forceps and a dissecting probe. Then, fill up the mold to its top and allow it to cool and solidify by putting it on top of a cooling platform. Section the embryos at uniform thickness from 8 to 15 µm. Prior to sectioning, preclean slides to remove the oily film which is even present on commercial slides that are identified as precleaned (see Note 5). Place a uniform number on all slides. We put five sections in a row and two rows on a slide. This facilitates ease in accounting for all of the sections. If a section is lost, leave a space where that section should be. Once the embryo is completely sectioned, process the slides for staining (see Note 6) and mounting. Screen the slides for the beginning and end of the sections containing brain. Circle every other section with a fine-point permanent marker. It helps to use a different colored marker to indicate the beginning and end sections from the color used to mark the interim sections. To measure using a computerized image analysis system, you will need a high-resolution camera connected to a high-quality compound microscope and computer with at least 16 MB RAM and abundant ROM. We use a MT1-CCD72, B & W camera connected to a Nikon Optiphot-2 (Nikon/Image Systems Inc., Columbia, MD) via a trinocular head and transmitting a split signal to an Ikegami monitor (Pike Technologies, Madison, WI) and a Power Mac 7500/100 (Apple Computer, Cupertino, CA). The monitor is an extra output, allowing us to view large images while the computer is being used. It is wise to setup the computer with a backup system for both the images and spreadsheet of data as soon as

Analysis of Brain Growth

13.

14.

15.

16.

171

you begin. A Zip drive or Jazz drive are good ways to save the data, or a local network hard drive if you have access. A convenient program, NIH Image, can be downloaded for free from the Internet. At the moment, this program is only available for a Macintosh. However, Sigma Plot (Jandel Scientific, San Rafael, CA), a commercial enterprise, has one similar that is available for both Macintosh and PC. Calibrate the computer for the objective and camera relay lens used for your measurements. Make sure the numbers make sense. A good way to check the output of your measurements is to measure a square of graph paper. You should also calibrate a mechanical polarizing planimeter the same way. Measure each section three times and only keep those measurements that are within 2–3% of each other. If you have several people measuring, which is most likely on a large project, make sure that their measurements agree with each other within 2–3% and check their results frequently. Weekly checks will save a lot of aggravation when you are trying to end a project. If your project requires many morphometric measurements, you will need to dedicate a computer to that purpose. Set up the spreadsheet to calculate the volume from the sections. Remember, you need to account for the total number of sections as well as the thickness of the sections. Be sure that all of the units are consistent; that is, if your volume is in cubic micrometers, be absolutely certain that the thickness is in micrometers. The formula for determining the volume is (Sum of the area) × (Section factor) × (Thickness of the sections). If you are using a mechanical planimeter or computerized planimeter that requires you to measure images projected by the microscope, you will need to divide the value obtained above by the magnification factor cubed (see Notes 7 and 8). Scan in the images to be measured into the computer, then identify regions that need measuring by using morphological markers. This is easy to do in an early chick brain because there are defined morphological boundaries between the first three vesicles and the next five vesicles. The boundaries are the telencephalomesencephalic fissure, the mesencephalorhombencephalic fissure, and the rostral edge of the otocyst. The program allows you to draw a straight line across from the indentation on the dorsal surface to that on the ventral surface. Several studies can be used as resources for morphological boundaries (3,9). Trace the outside perimeter of the brain and the inside. Then, subtract the inside from the outside. The outside tracing will give the total area of that brain section; the inside tracing will yield the total area of the cavity for that section. The difference between the two will be the total tissue area for that section (see Fig. 1A). Determine the cell number of a particular portion of the brain (e.g., the midbrain) by counting nonrandomized quadrats of the midbrain throughout its entire rostral to caudal extent. It is important to include a cell count from the rostral, mid, and caudal portions of the embryonic brain because it has been well documented for the chick embryo that rates of cell proliferation differ for these areas and for different regions of the brain; that is, the rates are not identical for the forebrain and midbrain (10–12). In other tissues, a nonrandomized sampling might be required because the density differs in various regions. To assure that you are not counting the same cells twice you need to skip enough sections to account for the thickness of a cell (see Note 9). Place an ocular reticle into one eyepiece of a compound microscope and count cells within each small unit of the reticle. The reticle will come with its dimensions (e.g., a 1 × 1-mm2 square divided into 10 units such that the area of each small unit is 0.1 mm2). Count the number of cells that lie in each square using a mechanical cell counter to keep track. (Actually, you will be calculating the number of nuclei, as it is nearly impossible to discern cell membranes of embryonic neuroepithelial cells.) Compare the cell numbers obtained for the different regions. If

172

Desmond and Haas

Fig. 1. The midsagittal section of a stage 24 chick embryo head in (A) illustrates the distinct borders separating the forebrain from the midbrain (1), the midbrain from the hindbrain (2), and the caudal end of the hindbrain (3). The large spaces surrounded by the neuroepithelium (ne) are the presumptive ventricles. The inner lining or ependyma of the neuroepithelium is labeled “i” and the outer border or apical surface is labeled “o.” The left and right photomicrographs in (B) feature lateral and dorsal views of the mesencephalon from a living, stage 24 chick embryo. In the lateral view (left view), the horizontal line = the anterior-posterior axis and the vertical line = the dorsal-ventral axis whereas a horizontal line drawn between the midpoints of the sides in the right view is the bilateral axis. These are the axes a, b, and c of an oblate spherioid used to calculate the volume.

Analysis of Brain Growth

16a.

16b.

16c.

16d.

17.

173

they are not significantly different (i.e., if they do not vary by 5%), use them all as an average. Extrapolate the total number of cells for the entire midbrain based on the relationship (Number of cells)/1 mm3 = Number of cells/(Midbrain volume in mm3). You need to have calculated the volume of the midbrain as previously described. Finally, a cautionary note is important here: Cell proliferation does not always lead to immediate growth in embryos; therefore, the measurement of cell proliferation rates is an inadequate indicator of growth. This is why cell numbers reported in the context of tissue volume is so important. Another way to determine the cell number of the embryonic brain is to determine the total DNA content of the neuroepithelium using the standard diphenylamine reaction (13–15) and converting this value to diploid equivalents by dividing this value by 2.5 pg, which is the amount of DNA reported in chick diploid cells (16). Unfortunately, this technique has the major disadvantage of destroying the embryonic brains because they are homogenized in the process, and the minor inconveniences of requiring total isolation of the neuroepithelium from surrounding mesenchyme and skin ectoderm plus estimating the proportion of cells that are not diploid. Such is the case because the cell cycle is asynchronous in the embryonic neuroepithelium so that a certain proportion of neuroepithelial cells will be haploid and tetraploid. If you assume that cells have the 4 N amount of DNA during all of mitosis (M), G2, and half of synthesis (S), the actual number of cell can be obtained with the formula: C = d/(1 + p) where d is the number of diploid equivalents and p is the proportion of cells expected to be 4 N (obtained by dividing the duration in hours of 1⁄2 S + G2 + M by the total generation time of the brain cells in hours at that stage (17). To determine the total DNA of a preweighed pooled sample of whole chick brains, (1) homogenize a known weight of tissue in p. saline with a hand-held Potter–Elvehjem homogenizer; (2) precipitate the macromolecules with cold 0.5N perchloric acid (PCA) followed by centrifugation (1290g for 15 min); (3) hydrolyze the nucleic acid (NA) with 0.5N PCA at 80°C for 40 min; (4) cool the tubes on ice for 5 min; (5) centrifuge and collect the supernatants for reaction with fresh diphenylamine reagent at 30°C overnight; (6) read the colored product formed at 595 and 700 nm visible light in a spectrophotometer; (7) determine the amount of DNA on a standard curve. To determine the number of diploid equivalents from the total DNA, calculate the amount of DNA per head by dividing the total amount of DNA by the number of heads in the pool. Then, divide this value by 2.5 pg, which is the mean of the highest and lowest values reported for chick diploid nuclei by Davidson et al. (18). If the tissues investing the brain are impossible to remove, then the above DNA concentration can be determined for a pool of whole heads from which the proportion of cells of nonbrain tissue is subtracted from total cell number. The proportion of brain and nonbrain cells can be determined by multiplying the cell density times the volume of that particular tissue. Cell density is determine by counting the number of nuclei in 100 squares of an ocular reticule in the dorsal and ventral regions of each of the brain vesicles included in the original brain sample excised and of regions of mesenchyme and epidermis adjacent to these brain regions. Volumes can be determined as described in step 15. Determine the mitotic index or density by counting colchicine-induced metaphases per 100 cells or volume, respectively. Prior to sacrificing embryos, treat embryos with 0.2 mL of colchicine (3.25 mg/mL); that is, the embryo receives 0.65 mg of the colchicine. This concentration has been worked out experimentally in our lab. Do not inject the colchicine into the yolk, but, rather, inject it right on top of the embryo as detailed in Note 10. Treat the embryos for 2–24 h. If more than 24 h, be sure to account for the added age of the embryo in your analysis. After sectioning the embryos, count the number of metaphase

174

Desmond and Haas

figures per total cells, as outlined in step 16. Then multiply this figure by 100. This is the mitotic index. Calculate the total number of metaphase cells per unit volume. This is the mitotic density. 18. Determine the DNA synthetic index and density by prelabeling the cells with either BrdUr or tritiated thymidine (3H-thymidine). For many reasons, the most notable of which is the cost of getting rid of radioactive waste and long exposure times required, we prefer the BrdUr method. However, it is possible to use both to get a precise measure of the actual generation time of the cells in the population (19). Inject the living embryo, as detailed in Note 10, with BrdUr (19) 2–6 h before sacrifice. Prepare sections as described and treat these sections with the primary antibody, anti-BrdUr, followed by a secondary antibody conjugated with horseradish peroxidase that will degrade the substrate diaminobenzidine, forming a brown–black product in the nucleus. Several kits are available for detection of BrDU. Although we have used the alkaline phosphatase–NBT, BCIP regime as well as the true blue substrate (3,3',5,5'-tetramethylbenzidine) by Kirkegaard and Perry Laboratories (Gaithersburg, MD), we find the kit by Calbiochem particularly reliable. We have also found the brown–black precipitate easier to detect against other staining. We use a very light hematoxylin counterstain so that nonlabeled cells can be detected. Once the cells are labeled, calculate the percentage of labeled cells and the volume to determine the DNA synthetic index and DNA synthetic density respectively. Be careful not to infer that DNA synthesis is an indication of mitosis. Synthesis and mitosis are two discretely separate phases of the cell cycle. 19. Label DNA using 3H-thymidine by injecting the embryos 1–6 h prior to sacrifice with 0.2 mL of 3H-thymidine having a high specific activity (i.e., 40 C/mmol). The high specific activity is to assure that an adequate signal will be generated by the nucleus. Thymidine is required because it has a very short tracking distance and, thus, it can be inferred that it is located within 0.2 µm from the place it is detected. Process the embryos for sectioning as already described. Once the sections are put on the slides, emulse them in liquid photographic emulsion (Kodak, Rochester, NY, NTBII). This must be done in a darkroom with only a red safelight on. Do not use the regular yellow darkroom safelights. Two to four hours prior to using the emulsion, place it in an incubator at 37°C or oven at 4°C. Gently shake the emulsion to see if it has liquefied. Then, take the following into the darkroom: a tissue flotation water bath preset to 48°C; preheated water for the bath; a heavy clamp to hold a plastic slide mailer tube upright in the water bath; test tube racks that can hold the slides once emulsed; a small container of distiller water; a red safelight; the premelted emulsion and slides to be emulsed. Add 7 mL of distilled water to the mailer, followed by 7 mL of emulsion. Close the cap of the mailer and gently invert it back and forth to mix. Be careful not to make bubbles that will interfere with the coating process. Place the mailer into the clamp and then put the apparatus into the water in the water bath. Be sure that the water covers at least two-thirds of the mailer height. Place two slides back to back and dip once into the diluted emulsion. Immediately pull apart and place vertically in separate spaces in the test tube rack for drying. Emulsing two slides together has the advantage of saving on emulsion by not coating the backside of the slides and cutting the time for emulsing in half. (A mailer full of emulsion will coat approximately 40 slides.) Once all of the slides have been emulsed and placed in a rack(s) to dry at room temperature (RT), place the rack(s) in a light-tight drawer or box to dry for 2–4 h. You may simply cover the racks with aluminum foil. Do not leave at RT for more than 4 h, as you risk getting excessive background. Now, transfer the dried, emulsed slides to slide boxes, label the box, and cover it with aluminum foil to prevent light leakage. Store the covered slide box at 4°C for 4–6 wk. After 3 wk, develop a test slide to see how much the emulsion has been exposed by the radioactivity.

Analysis of Brain Growth

175

20. Develop the slide(s) as follows. Have four containers for photographic chemicals. Use Coplin staining jars if only planning to develop a few slides; otherwise, use standard staining dishes with appropriate holders. The four containers will hold developer, distilled water (DW), and photographic fixer. In general, 200 mL of solution will process 100–200 slides before becoming exhausted and needing replacement. We prefer Kodak D-19 developer (Eastman Kodak, Rochester, NY) but many other labs use Dektol (Kodak) with excellent results. Whichever developer is used, be sure to make the concentration used for film. Set up the solutions with only a red safelight on and, once ready, open the slide box and place the emulsed slides into slide holders. Develop 5 min, rinse 5 min in DW, fix 5 min, and rinse for 5 min in DW. After all of the slides have been fixed, the slides may be exposed to regular room light. All of the emulsion should be removed from the slides because of the fixing process. Now the slides are ready to be stained and permanently mounted for viewing.

3.2. Morphometrics from Whole Embryos 1. Repeat steps 1–7 of the above procedure. Thus, the embryos will be fixed and stored in 70% EtOH. Begin with embryos that are between stages 20 and 28 of development. 2. Prepare an embryo holder for the head as follows. Place a circle of black construction paper in the bottom of a 60-mm Petri dish and fill the dish with paraffin. Once the paraffin in solidified, use a small metal spatula or rounded glass rods to make the appropriate size and shape depression to hold the head in its various orientations: lateral, dorsal, ventral, and frontal planes. Small glass rods made from small capillaries are helpful to provide weights in various locations on the embryo. To keep the rod from sticking to the embryo, siliconize it (Note 11). The objective is to hold the head in its perfect orientation so that a photograph can be taken. All of the measurements will be taken from the photographs. 3. Three axes of the midbrain can be measured: median anteroposterior, dorsoventral, and bilateral from identically magnified photographs (Fig. 1B). Using these measurements, the volume of the midbrain can be calculated using the mathematical formula for an oblate spheroid, (4/3) π abc, where a = c > b (5). We have shown that the upper and lower hemispheres are noncongruent; however, volume differences based on the midbrain having congruent hemispheres vs. non-congruent hemispheres are negligible. Differences in volumes based on a = c compared to those where a = c are also negligible. 4. Setting up the embryo for photographing in the four planes is time-consuming and requires tremendous patience. Not only is positioning the embryo tedious, but setting up reflected lighting without glares is also. Each position takes about 20–30 min to be set up. Thus, to get all of the planes photographed for one embryo will take 80–120 min. However, the payoff is tremendous when you think that it takes a minimum of 4–5 d to determine the same information from sections. A major use of this technique would be for pharmacological studies in which one wants to compare brain volumes of treated embryos with untreated embryos. Another use might be that of measuring volumes of the midbrain from embryos, which had been used for WISH.

4. Notes 1. Avoid dumping the egg residue, yolks, and whites excluding shells down the sink without abundant rinsing of the pipes with excess cold water. It is very easy to clog long segments of piping by running hot water into a line that still has egg residue. The best way to discard of the egg residue is to dump it down the toilet or freeze it in a plastic bag and discard with animal waste. The shells can be disposed of by placing in a plastic bag and dumping in the trash. Check with the animal waste committee of the institution about egg disposal.

176

Desmond and Haas

2. Hemode/histosol are both synthetic xylenes available from commercial vendors. They have the advantage of smelling less offensively but still do the same liver damage; therefore, it should be used in a chemical hood. 3. Be sure that the timing of the solutions containing either pure paraffin or partial paraffin are based on when the paraffin is melted. Be very careful to have the oven or hot block just 3–5°C above the melting temperature of the paraffin so as not to “cook” the embryos and lose all morphology. Finally, only leave the embryos in the melted paraffin for 1 h. However, you may store them in solidified paraffin at 4°C for as long as necessary. 4. Embedding molds and rings: The easiest system for embedding is to use stainless-steel base molds on top of which sits a plastic ring. Another commonly used system is the peelaway plastic molds. Use of these molds necessitates placing a paper identification tag into the top of the liquid paraffin before it solidifies. Both are readily available at either Fisher or VWR Scientific (Atlanta, GA). 5. Cleaning of slides: Place slides in slide carriers and transfer for 5 min each through the following solutions: three changes of sodium dichromate–sulfuric acid cleaning solution; three changes of distilled water; three changes of absolute alcohol. Air-dry by placing a beaker over the slide carrier with an inch gap at the bottom to allow airflow. Be sure to keep them covered or the slides will get dusty. Once dry, place the slides in a slide box for storage. 6. Preliminary to staining, deparaffinize the sections by placing them in a slide box, then placing the box into an oven at the melting temperature of the paraffin for no longer than 30 min. Remove to room temperature and allow to cool. Transfer the slides to a slide carrier and begin the chemical hydration process: (a) pass the slides through the following chemicals for 5-min intervals: three changes of histosol/hemode; two changes of absolute ethanol; two changes of 95% ethanol; two changes of 70% ethanol; one change of DW; (b) stain the slides 5–8 min in Harris’ Hematoxylin that has been prefiltered; (c) rinse in running tap water until no color leeches out into the water bath; (d) dip 10–20 s in 1% acid alcohol; in running water bath for 15 s; in saturated lithium carbonate for 15 s; in running tap water for 15 s; (e) stain in alcoholic Eosin for 5–10 min; (f) rinse through three changes of 95% ethanol, 15–20 dips each; (g) transfer through two changes of absolute ethanol, 5 min each; finally, three changes of hemode for 5 min each. Mount with permount or DPX. Allow to dry thoroughly. 7. The magnification factor is determined by projecting the focused image of a stage micrometer onto a piece of paper and tracing the lines. It can also be photographed. Be sure that it is projected at the same magnification as the sections. This is the magnified image. Measure several (six to eight) of some known increment of the traced micrometer with a good ruler. Be sure to measure the same side of the lines, not the middle. Take the average of these measurements and you have the average magnified distance. Divide this value by the value given for the same increment on the stage micrometer. Often the smallest increment is 0.01 mm. If the magnified image of the smallest increment is 4.5 mm, then the magnification factor would be 4.5 divided by 0.01 mm, or 450. 8. The details may vary for your particular computer setup, but a stepwise procedure for capturing and measuring images with a Mac can be obtained from the Internet together with the NIH image. 9. Neuroepithelial cells are elliptical in shape in early embryos. Thus, we measured the short and long axes of several cells. We took the average of each and then calculated the volume of the cell. A best approximation of the thickness of a cell is 25 µm based on the longitudinal axis. Thus, if the sections are cut at 10 µm, at least 2.5 sections need to be skipped in between the ones used for cell counts. A common comparator for cell size is to remember that a human red blood cell is 8 µm in diameter.

Analysis of Brain Growth

177

10. Candle the fertilized egg and, with a pencil, draw a circle on the shell that indicates the edge of the blastodisk. Swab this inscribed area with an alcohol sponge, then make a small hole (0.5 mm) in the shell using a sanding disk attached to a Dremel drill (Dremel, Racine, WI). Insert the end of a 18-gage needle that is attached to a 1.0-mL sterile syringe containing the colchicine solution. Slowly inject the colchicine. Remove the needle and seal the hole with 3M transparent tape. 11. Dip glass probes for 2 s into a 1:100 diluted Prosil solution (Thomas Scientific) followed by 2-s dips in three changes of DDW. Place the wet probes onto the right half of an opened Kimwipe that has been placed on the bottom of a plastic container. Cover the probes with the left half and place more siliconized probes on the top of the left half. Add another opened Kimwipe for more tips. Cover the plastic container when finished to keep tips clean.

Acknowledgment The authors thank Antone G. Jacobson for introducing one of us (MED) to the art of embryo microsurgery and the intricacies of morphometric analyses. Both authors thank the many graduate and undergraduate students, especially Bernard Martin, who have developed techniques for working with embryos. This work is supported by grants from the NIH HD18143 and NS24136. References 1. Desmond, M. E. and Jacobson, A. G. (1977) Embryonic brain enlargement requires cerebrospinal fluid pressure. Dev. Biol. 57, 188–198. 2. Weibel, E. R. (1979) Stereological Methods. Vol. 1. Practical Methods for Biological Morphometry. Academic, New York. 3. Desmond, M. E. and O’Rahilly, R. (1981) The growth of the human brain during the embryonic period proper. 1. Linear axes. Anat. Embryol. 162, 137–151. 4. Summerbell, D. (1976) A descriptive study of the rate of elongation and differentiation of the skeleton of the developing chick wing. J. Embryol. Exp. Morphol. 35, 241–260. 5. Desmond, M. E., New, M. S., Martin, B. G., and Fleischman, W. M. (1990) A rapid reliable calculation of brain expansion in living chick embryos to reflect fluid transport across the neuroepithelium. Anat. Rec. 226, 34A. 6. Gibson, K. D., Segen, B. J., and Doller, H. J. (1979) B-D-xylosides cause abnormalities of growth and development in chick embryos. Nature 273, 151–157. 7. Desmond, M. E. and Field, M. C. (1990) Bubble-like minispheres of avian embryonic neuroepithelium: a model for studying fluid transport. Soc. Neurosci. Abstr. 16, 1150A. 8. Hamburger, V. and Hamilton, H. (1951) A series of normal stages in the development of the chick embryo. J. Morphol. 88, 49–92. 9. Pacheco, M. A., Marks, R. W., Schoenwolf, G. C., and Desmond, M. E. (1986) Quantification of the initial phases of rapid brain enlargement in the chick embryo. Am. J. Anat. 175, 403–411. 10. Kallen, B. (1961) Studies on cell proliferation in the brain of chick embryos with special reference to the mesenchephalon. Z. Anat. Entwicklungsgesch. 122, 388–401. 11. Cowan, W., Martin, A., and Wenger, E. (1968) Mitotic patterns in the optic tectum of the chick during normal development and after early removal of the optic vesicle. J. Exp. Zool. 169, 71–92. 12. Wilson, D. B. (1973) Chronological changes in the cell cycle of chick neuroepithelial cells. J. Embryol. Exp. Morphol. 29, 745–751. 13. Burton, K. (1956) A study of the conditions and mechanism of the diphenylamine reaction for the colorimetric estimation of deoxyribonucleic acid. Biochem. J. 62, 315–323.

178

Desmond and Haas

14. Giles, K. and Myers, A. (1965) An improved diphenylamine method for the estimation of deoxyribonucleic acid. Nature 206, 93. 15. Dische, Z. (1930) Ueber einige neue characteristische farbreaktionen der thymonukleinsaure und eine mikromethode zur bestimmung derselben in tierischen organen mit hilfe dieser reaktionen. Mikrochemie 8, 4–32. 16. Desmond, M. E. (1985) Reduced number of brain cells in so-called neural overgrowth. Anat. Rec. 212, 195–198. 17. Wilson, D. B. (1974) The cell cycle of ventricular cells in the overgrown optic tectum. Brain Res. 69, 41–48. 18. Davidson, J., Leslie, I., Smellie, R., and Thomson, R. (1950) Chemical changes in the developing chick embryo related to the desoxyribonucleic acid content of the nucleus. Biochem. J. 47, XL. 19. Hyatt, G. A. and Beebe, D. C. (1992) Use of a double-label method to detect rapid change in the rate of cell proliferation. J. Histochem. Cytochem. 40, 619–627.

Neuroepithelial Minispheres

179

19 Isolation of Neuroepithelium and Formation of Minispheres Mary E. Desmond and Marcia C. Field 1. Introduction The brain of the early chick embryo initially grows by expansion of its fluid-filled cavity much as a balloon expands (1–3). Such expansion occurs rapidly over 3–4 d, beginning at around 43 h of development (1,3), and is under the direct control of internal fluid pressure, as shown from brain intubation experiments (1,4–8). Intraluminal pressure is generated by the accumulation of fluid within a closed system and promotes expansion of the brain. The source of the fluid within the early neural tube still remains unknown, although Weiss (9) reported that the neuroepithelium is secretory as soon as it forms a tube. Weiss’s experiments are only suggestive, however, and need to be repeated using contemporary methods in order to establish for certain whether the neuroepithelium is secretory, and if it is, at what stage it first becomes so. In addition, no information exists about how fluid crosses the neuroepithelium during this early expansion period. Although several reports exist summarizing the secretory nature of the choroid plexus, the choroid plexus does not develop until much later in embryogenesis. Our laboratory has adopted a technique devised by Stern et al. (10), in which he cultured ministrips of embryonic epiblast that became fluid-filled spherical vesicles. We cultured explants from whole-head segments and strips of either neuroepithelium or skin ectoderm (11). Although we found that the whole-head segments sorted out into two primary tissues and configurations (a mesenchymal glob and ectodermal fluidfilled sphere [minispheres], we were unable to distinguish those spheres that were purely epiblast from those that were purely neuroepithelium. To address this question, we developed a microsurgery technique that produces pure neuroepithelium and epiblast strips. Although such a technique requires some skill in microsurgery, it provides pure neuroepithelial minispheres that can be cultured long term and used for various physiological and pharmacological studies involving assessment of fluid transport across the neuroepithelium. 2. Materials 1. Crude trypsin: Type III from Sigma (St. Louis, MO), catalog #T8253. Prepare immediately or within the hour of use. From: Methods in Molecular Biology, Vol. 136: Developmental Biology Protocols, Vol. II Edited by: R. S. Tuan and C. W. Lo © Humana Press Inc., Totowa, NJ

179

180

Desmond and Field

2. Tungsten needles: Prepare these needles from the finest tungsten wire available (i.e., that used in electric light filaments). The outside diameter of this wire should be no greater than 0.2 mm. Cut a 3-in. segment from the wire supply and thread it through a 23-gage needle, allowing 5–10 mm to extend from the sharp beveled end of the needle. Taper one end of a 6- to 8-in. length of wood dowel to fit into the plastic end of the needle used to attach the needle to a syringe. Be sure that the dowel is at least 6 in long so that it is comfortable as a handle for doing fine manipulations. Insert the tapered end into the dowel, wrap the excess tungsten wire around the wood dowel near the junction with the plastic end of the needle, and then wrap the junction site with electrical tape. Electrical tape is preferable to other tapes because it has more stretch and makes a tight wrapping around the junction. We like to use enough tape to make a comfortable holding position along the dowel. This means that the thickness where you grasp the dowel is like that of a comfortable pen. Once the wire is secured into the dowel, sharpen the free end of the wire by inserting it in and out of a solution of saturated sodium nitrite. In order for ionic displacement to occur, either the wire or the sodium nitrite needs to be very hot. It is easier and faster to sharpen the wire by heating it as it is inserted. This is done by applying an electrode to the metal portion of the needle while attaching the other part of the electrode to a reference carborundum rod from a D-cell battery. The easiest source for the electrical reference and sample electrodes is to use a power supply for an external microscope light (e.g., Bausch and Lomb, Rochester, NY, 125 V). Make the reference electrode by attaching copper wire around the corborundum rod and bending the free end of the wire into a hook that can be hung at the top of a 100-mL straight-sided glass beaker holding the saturated sodium nitrite. Once the needle is attached to the sample electrode and the reference carborundum rod is inserted into the solution, turn on the power of the power supply to 12 V and insert the last 2 mm of the tungsten wire in and out of the sodium nitrite solution. Bubbles will appear at the end of the wire. Continue doing this until you get the thinness of point that you desire. Caution: Do not simply leave the end segment of the wire in the solution while current passes through it, because the wire will be uniformly displaced and become a thin filament too flexible for using as a cutting device. By inserting the end of the wire in and out the solution, a tapered end like a pencil point is formed. The saturated solution of sodium nitrite can be used over and over for years. Although the preparation of these tungsten needles takes some time, the results are cutting needles that keep their sharpness for several operations and never break, in contrast to glass needles that dull more quickly and easily break. To store these needles, insert the base of the wood handle into a 4 × 5 × 2 in. piece of styrofoam and cover the entire construct with an inverted plastic beaker. Another storage method involves putting the wood base into a 15-mL plastic conical tube, covering the rest of the needle with another conical tube, and taping the two tubes at the junction. This forms a protector similar to a cigar storage cylinder. To keep the tungsten needle from moving back and forth, place some styrofoam in the bottom of the protective tube that will hold the base securely. Finally, the tungsten can be sharpened by boiling the solution of saturated sodium nitrite. Although both methods can be dangerous, this one emits noxious fumes and takes at least 10 time longer than the electrical method. The solution needs to be in a porcelain crucible and the heat needs to be at a constantly high temperature. It is essential to use in a chemical hood and heatprotective gloves. 3. Hanks balanced salt solution, modified (10X) Catalog #19-101 (without NaHCO3 from Flow Laboratories, Inc. (Rockville, MD, USA). 4. Enriched balanced salt solution: 94 mL Medium 199 plus 2.2 g NaHCO3 plus 5 mL fetal bovine serum or 5 mL Chang C medium plus 1 mL penicillin (25 units/mL), streptomycin

Neuroepithelial Minispheres

181

(0.25 mg/mL). Penicillin–streptomycin is from Gibco-BRL, Gaithersburg, MD, catalog # 600-5140-AG. 5. Medium 199 (10X) is from Gibco-BRL, catalog #330-1180AG (100 mL). 6. Chang C medium: Catalog #T104 from Irvine Scientific (Santa Ana, CA). Developed for the primary culture of human amniotic fluid, this fluid contains salts, glucose, amino acids, polypeptides, vitamins, ribonucleotides, deoxyribonucleotides, sodium pyruvate, steroid hormones, and bovine serum proteins. 7. Miniwell plates with 24 wells.

3. Methods 3.1. Minisphere Formation from Isolated Head Segments 1. Repeat steps 1–4 inclusive of the methods in Chapter 18. This technique can be applied to chick embryos of HH stages 7–15. Use sterile techniques throughout (see Note 1). Aspirate off the p. saline using a Pasteur pipet and add 2–3 mL of 1% crude trypsin in p. saline. Place the Petri dish on an ice bath, being careful that the water of the bath does not get into the Petri dish. A standard glass embryo culture dish makes a nice container for the ice bath. Incubate at 4°C for 20 min, followed by five rinses in Hanks’ balanced salt solution (BSS) maintained at room temperature (17–18°C). 2. Flatten the embryo by aspirating around the perimeter of the blastodisk with a Pasteur pipet. Hold the head of the embryo in position by placing the tines of a pair of straight tipped forceps on either side of the head. Draw the point of a sharp tungsten needle along the boundary between the forebrain and midbrain, midbrain and hindbrain and at the caudal border of the hindbrain (cf. Fig. 1 in Chapter 18). 3. Transfer each of the head isolates into separate miniwells of a 24-well culture plate containing 1 mL of BSS plus 10% calf serum plus 1% penicillin and streptomycin. Several (four to five) head isolates may be cultured in one well. 4. Place the miniwell cultures in a 37°C humidified incubator, biological oxygen demand. Check daily for the presence of minispheres. The minispheres take a minimum of 24 h to form and will continue living for as long as 2 wk if the medium is changed daily. The viable minispheres will appear transparent and under tension. The volume can be monitored on the living spheres by measuring the diameter of the sphere using an ocular micrometer and applying the formula, V = 4°π = 3.1416 and r = radius = 1⁄2 d (see Note 2). 5. For a permanent record, the minispheres can be photographed through a dissecting microscope using transmitted light, then printing the photographs. The diameter can be easily measured from the photomicrographs. However, now the volume must be corrected for the magnification factor in the same manner as described in Note 7 of Chapter 18.

3.2. Preparation of Pure Neuroepithelial and Skin Ectoderm Explants 1. Repeat steps 1–4 of the method described in Chapter 18 and step 1 above using stage 5 to stage 15 embryos. 2. Prepare the embryo explant by cutting away all of the tissue lateral, anterior and posterior to the embryonic axis. First, using a sharp tungsten needle, transect the top edge of the embryo axis by cutting an arc at the border of the top of the head, extending laterally along the top edge of the skin ectoderm. Then, on both sides of the embryo axis proper, cut away the lateral tissues by making a longitudinal cut parallel to the embryo axis at the border of the skin ectoderm with a less differentiated epiblast (lateral cut). Next, make a longitudinal cut parallel to the embryo axis at the border of the neuroepithelium with the skin ectoderm (medial cut). Finally, make a posterior cut across the longitudinal segment of the embryo axis at the level of Hensen’s node. There are now three distinct segments of

182

Desmond and Field

tissue present in the Petri dish: two fairly identical rectangles of skin ectoderm atop endomesoderm plus the embryo axis proper, also consisting of neuroectoderm (topmost layer), mesoderm (middle layer), and endoderm (innermost layer). 3. Separate the topmost layer of neuroectoderm of the embryonic axis from its underlying endomesoderm layers by holding the endomesoderm down with a dissecting probe and gently pulling away the neuroectoderm with another set of forceps. Once the neuroectoderm of the entire length of the neural plate has been separated from its underlying tissues, cut away the underlying tissues using a tungsten needle. Then, place the neural plate down on the Petri dish surface and transect it from the rest of the embryo axis at the superior level of Hensen’s node. Transfer this pure neuroectoderm to a miniwell containing 1 mL of enriched BSS by using a Pasteur pipet (see Fig. 1A–F). 4. Similarly, separate the top layer of skin ectoderm from each of the previously prepared rectangles of embryonic tissue lateral to the neural tube axis by holding the endomesoderm down onto the surface of the Petri dish with a dissecting probe while pulling away the skin ectoderm. Then, place these two rectangles of skin ectoderm into a miniwell containing enriched medium. 5. Monitor the cultures of pure neuroepithelial and skin ectoderm explants daily for lack of contamination, viability of the tissues, and formation of minispheres. The skin explants will most likely form minispheres after 1 d in culture, whereas the neuroepithelial explants will take a minimum of 2 d. Minispheres produced from each of the tissues can be cultured long term (i.e., at least 2 wk and no doubt longer if the medium is changed every other day). Volumes can be determined from these minispheres exactly as described in steps 4 and 5 of Subheading 3.1.

4. Notes 1. Although embryo cultures are more forgiving than mammalian cell culture, sterile technique needs to be quite rigidly followed. The following procedures should be followed as a minimum. Ideally, work in a laminar-flow hood. Whereever you work, swab the working surface with 70% ethanol before working. Have the microscope positioned close to the incubator. Use sterile glassware or plasticware throughout. Sterilize the tungsten needles in between touching the embryo by dipping them in soapy water, followed by distilled water (DW), followed by ethanol, followed by flaming in a low Bunsen burner flame or alcohol flame. We use 100-mL glass beakers for our sterilizing solutions. Sterilize all instruments using this method. This is very important. The reason we use soapy water is because chick embryo tissue is very sticky and does not easily come off with just water. Keep your nose away from the operating site (i.e., use the power of your eyes to see and do not put your face down into the culture). Wear a mask and gloves, as the procedure takes 20–30 min. 2. Because the micrometer is inside the ocular, it will not change size when different objectives are used. However, the diameter of the minisphere will increase in length as the sphere is magnified by a higher-power objective. Therefore, to determine the amount of magnification, multiply the power of the objective by the power of the ocular, then divide the diameter length by this number. Then, calculate the actual volume by the formula noted in item 4 of Subheading 3.1.

Acknowledgment Both authors thank the many graduate and undergraduate students who have developed techniques for working with living chick embryos. This work is supported by grants from the NIH HD18143 and NS 24136.

Neuroepithelial Minispheres

183

Fig. 1. This series of photomicrographs illustrate the formation of pure neuroepithelial minispheres (F). The embryo in (A) is a stage 5 embryo, dorsal side up, that has been excised off the yolk, ready for isolation of the neuroectoderm from the underlying endomesoderm. Hensen’s node is marked by the asterisk. View (B) shows the absence of the skin ectoderm lateral to the neural plate as well as the neuroectoderm pulled away from the underlying endomesoderm. In (C), the neuroectoderm (right side of the forked tissue) is clearly identified, apart form the underlying endomesoderm on the left. View (D) shows the isolated neural ectoderm before it has been cut away from the embryonic axis at the superior level of Hensen’s node. This is the same piece of tissue shown on the rights in view (C). Once the isolated neuroectoderm is put into liquid culture, it curls up as shown in (E) and after 2 d, it sorts out into a hollow neuroepithelial minisphere (ms in F) with associated extracellular material (glob). The bar represents 2.50 mm for (A)—(D) and 2.54 mm for (E) and (F). SE = skin ectoderm; NE = neuroepithelium (neural ectoderm); E = explant; ms = minisphere; asterisk = Hensen’s node.

184

Desmond and Field

References 1. Desmond, M. E. and Jacobson, A. J. (1977) Embryonic brain enlargement requires cerebrospinal fluid pressure. Dev. Biol. 57, 188–198. 2. Desmond, M. E. (1985) Reduced number of brain cells in so-called neural overgrowth. Anat. Rec. 212, 195–198. 3. Pacheco, M. A., Marks, R. W., Schoenwolf, G. C., and Desmond, M. E. (1985) Quantification of the initial phases of rapid brain enlargement in the chick embryo. Am. J. Anat. 175, 403–411. 4. Coulombre, A. J. and Coulombre, J. L. (1956) The role of intraocular pressure in the development of the chick eye. 1. Control of eye size. J. Exp. Zool. 133, 211–225. 5. Coulombre, A. J. and Coulombre, J. L. (1958) The role of intraocular pressure in the development of the chick eye. IV. Corneal curvature. A.M.A. Arch. Ophthalmon. 59, 502–506. 6. Coulombre, A. J. and Coulombre, J. L. (1958) The role of mechanical factors in the brain morphogenesis. Anat. Rec. 130, 289–290. 7. Jelinek, R. and Pexieder, T. (1968) The pressure of encephalic fluid in chick embryos between the 2nd and 6th day of incubation. Physiol. Bohemoslav. 17, 297–305. 8. Jelinek, R. and Pexieder, T. (1970) Pressure of the CSF and the morphogenesis of the CNS. Folia Morphol. 18, 102–110. 9. Weiss, P. (1934) Secretory activity of the inner layer of the embryonic mid-brain of the chick as revealed by tissue culture. Anat. Rec. 58, 299–302. 10. Stern, C., Manning, S., and Gillespie, J. (1985) Fluid transport across the epiblast of the early chick embryo. J. Embryol. Exp. Morphol. 88, 365–384. 11. Desmond, M. E. and Field, M. C. (1990) Bubble-like minispheres of avian embryonic neuroepithelium: A model for studying fluid transport. Soc. Neurosci. Abstr. 16, 1150A.

Examining Placentation in the Mouse

185

20 Examination of Normal and Abnormal Placentation in the Mouse Michael R. Blackburn 1. Introduction Placental development, also known as placentation, is orchestrated by precise molecular and cellular interactions between extraembryonic cells and cells of the pregnant uterus. The major extraembryonic cell type of the placenta is the trophoblast cell, which arises from the trophectoderm cell lineage (1). Trophoblast cells make the physical connection between the embryo and the maternal environment and play important roles in the implantation process and placental function (2). In mice, trophoblast cells lead the embryonic invasion into the uterine stroma to achieve implantation, and contribute to the formation of a temporary yolk-sac placenta (choriovetelline placenta) during early postimplantation stages. Trophoblast cells eventually provide the major extraembryonic cell lineage that contribute to the formation of the mature chorioallantoic placenta (3). In the chorioallantoic placenta, the fetal and maternal circulatory systems are brought into close apposition to allow for a range of physiological interactions that are critical for fetal growth and development. Abnormal placentation is often associated with early embryonic mortality (4–6) and can lead to serious pregnancy disorders such as preeclampsia (7). The importance of placental development in the mouse has recently been highlighted by the targeted disruption of genes that lead to placental phenotypes that, in turn, cause embryolethality (4–6). An example of this is the targeted disruption of the mash-2 gene, which encodes a basic helix–loop–helix transcription factor that is expressed in developing trophoblast cells (4). Absence of this transcription factor results in the absence of a population of trophoblast cells known as spongiotrophoblast cells, which in turn, results in abnormal placentation and the death of the embryo by mid-gestation. In addition, genes that have not previously been known to serve a role during placentation are found to be important when their disruption leads to a placental phenotype (6). It is therefore critical that researchers closely investigate the placenta when assessing the phenotypic outcome of gain of function or loss of function mutations in transgenic mice. Typically, extraembryonic tissues are discarded in the process of dissecting the embryo or fetus from the uterus for analysis. In this chapter, a procedure will be From: Methods in Molecular Biology, Vol. 136: Developmental Biology Protocols, Vol. II Edited by: R. S. Tuan and C. W. Lo © Humana Press Inc., Totowa, NJ

185

186

Blackburn

described for the processing and analysis of gestation sites with a particular emphasis on analyzing intact extraembryonic tissues for the analysis of maternal–embryonal cellular interactions. In addition to describing the techniques for processing gestation sites for histopathological analysis, a technique for studying giant trophoblast cell populations will be presented. 2. Materials 2.1. Fixation and Embedding of Gestation Sites 1. Phosphate buffered saline (PBS): Make 10X stock solution: Start with 800 mL H2O, add 80 g NaCl, 2 g KCl, 14.4 g NaHPO4 · 7H2O, and 2.4 g KH2PO4. pH to 7.4 using HCl. Bring volume to 1 L with H2O, and autoclave. The week of use, make 1X working buffer and autoclave. 2. 2M NaOH. 8 g NaOH into 100 mL H2O. 3. 2M HCl. 16.5 mL 12.1N HCl into 83.5 mL H2O. 4. 4% Paraformaldehyde (PFA) in PBS (see Note 1). Heat 90 mL H2O and 100 µL 2M sodium hydroxide to 50°C in a microwave oven. Add 4 g PFA (Fisher Scientific, Pittsburgh, PA) avoiding inhalation (wear mask) and stir in a hood until the PFA is dissolved. Add 10 mL of 10X PBS, mix well, and then filter solution through Whatman filter paper. Let the solution cool to room temperature and pH to 7.4 by adding 100 µL 2M HCl. Store tightly sealed at 4°C and use for up to 1 wk. 5. 20-mL glass scintillation vials (Kimball [Kimble/Kontes, Vineland, NJ]). 6. Graded ethanols. 7. Histoclear (National Diagnostics, Atlanta, GA) or Xylenes (Fisher). 8. Paraffin (Parablast, Fisher). 9. Embedding molds (Fisher). 10. Positively charged microscope slides, cover glasses (Fisher). 11. Slide-staining racks and dishes (Baxter, Muskegon, MI). 12. Fine scissors, blunt forceps, and a small hemostat (Thomas Scientific, Swedesboro, NJ). 13. 35-mm Plastic Petri dishes (Fisher).

2.2. Monitoring Placentation 1. Hematoxylin and Eosin Staining Kit (Shandon, Inc., Pittsburgh, PA). 2. Paramount mounting media (Fisher). 3. Hoechst, bisbenzimide (Sigma, St. Louis, MO). Prepare a 10-mg/mL stock solution by dissolving 100 mg bisbenzimide in 10 mL dimethylsulfoxide (DMSO, Sigma). Aliquot into 500 µL samples in 1.5-mL microfuge tubes, wrap tubes in foil, and store at –70°C until needed. 4. Methyl salicylate (Sigma). 5. Canada balsam (Sigma). Prepare Canada balsam mounting solution by mixing 5 g Canada balsam with 10 mL methyl salicylate. Store at room temperature in a light-protected container such as a bottle wrapped in foil.

3. Methods 3.1. Timed Pregnancies and Dissection of Gestation Sites 1. Mate two or three females (at least 7 wk of age), with one male (at least 7 wk of age) (see Note 2). Place females into the male’s cage at 16:00 and check for the presence of a vaginal plug the following morning (see Note 3). Detection of vaginal plug designates 0.5 d post coitum (dpc) (see Note 4). Keep plugged females caged apart from males.

Examining Placentation in the Mouse

187

2. At the desired stage of pregnancy, sacrifice mothers by cervical dislocation or carbon monoxide asphyxiation. Place sacrificed animal on its back and drench animal with 80% ethanol. 3. Lift the skin at the base of the abdomen (between hind legs) with blunt forceps and cut through the skin using dissecting scissors, revealing the abdominal muscle wall. Next, grasp the muscle wall, lift, and cut a hole in the muscle wall into the abdominal cavity. Continue to lift the muscle wall and cut through both the muscle and skin up both sides of the abdomen and resect away the muscle and skin to fully expose the abdominal cavity. Take care not to cut any internal organs or the uterus. The uterus can be located by lifting the intestines and viscera up and forward, out of the way. Locate the cervix at the base of the abdomen between the hind legs, grasp it with forceps, and cut between the cervix and vagina. Lift the cervix and cut along the underside of both uterine horns, cutting through the mesentery and uterine artery on the dorsal side of each uterine horn without cutting through the myometrium of the uterus itself. Remove the uterus by cutting between the ovary and the uterus. Place the intact uterus containing gestation sites into a 100-mm Petri dish containing ice-cold PBS. 4. Using forceps and scissors, trim away excess fat and vessels from the mesometrial side of the uterus (see Note 5). Attach a small hemostat to the cervix to help stabilize the uterus. Next, grasp the end of a uterine horn, stretch it slightly, and cut between each gestation site (see Fig. 1A). This will generate individual gestation sites contained in the myometrium (see Note 6). Rinse separated gestation sites in fresh ice-cold PBS before transferring to fixative.

3.2. Tissue Fixation, Embedding, and Viewing 1. Place gestation sites into glass scintillation vials containing ice-cold 4% PFA in PBS (approx 20 mL). The number of gestation sites in each vial will depend on the stage of gestation (see Note 7). Fix gestation sites overnight at 4°C, with gentle rocking. 2. Wash gestation sites 2 × 30 min (see Note 8) in PBS, rocking at 4°C. 3. Dehydrate samples by incubating for 2 × 30 min, rocking at 4°C in a mixture of 30% ethanol in PBS, 50% ethanol in PBS, 70% ethanol in H2O, 85% ethanol in H2O, a mixture of 95% ethanol in H2O, and 100% ethanol. Gestation sites can be stored at this stage at –20°C or cleared and embedded. 4. Clear gestation sites by incubating 20 min, rocking at room temperature in a 1:1 mixture of ethanol and histoclear (or xylenes; see Note 9), followed by 2 × 20 min in 100% histoclear, rocking at room temperature. 5. Infiltrate gestation sites with paraffin by incubating 20 min (see Note 9) at 56°C in a 1:1 mixture of histoclear and paraffin, followed by 3 × 30 min at 57°C in 100% paraffin. 6. Embedding sections: Warm paraffin wax ahead of time by placing in a beaker overnight at 58°C. Do not overheat wax and discard heated wax after 2–3 d. 7. Fill embedding mold with melted paraffin. Using forceps, heated slightly over an ethanol flame, immediately remove a single gestation site and place in the mold. Using heated forceps, orient the gestation site such that the lumen of the uterus is facing the bottom of the mold (like a football standing on end, see Fig. 1B and Note 10). Do not allow the formation of air bubbles. If the gestation site is not oriented properly, remelt the paraffin block at 58°C and re-embed. 8. Allow blocks to cool to room temperature. Gestation sites in the blocks can be stored at room temperature indefinitely.

188

Blackburn

Fig. 1. Dissection and orientation of murine gestation sites for optimal visualization of extraembryonic membranes and placenta. (A) Diagram of uterine horns on approx 8.5 dpc. Gestation sites (gs) are defined as the developing embryo and its extraembryonic membranes contained within the uterus. To obtain an individual gs cut between decidualized areas of the uterus as depicted with lines. cv, cervix. (B) When embedding, orient the gs standing on end, with the uterine lumen facing down. (C) When sectioning, initial cuts are through the undecidualized uterus (a), then through decidualized uterine tissue devoid of embryonic of extraembryonic tissue (b). Begin collecting sections once the implantation chamber becomes evident (c): e, embryo; epc, ectoplacental cone; ul, uterine lumen; my, myometrium; m, mesometrial deciduum; a, antimesometrial deciduum. 9. Sectioning. Mount the embedded gestation site on a wooden block for sectioning (or comparable mounting apparatus). The bottom of the embedded block should be facing out (see Fig. 1B,C). 10. Using a rotary microtome, section through the gestation site and start collecting sections 5 µm in thickness when you enter the implantation chamber (see Fig. 1B,C). Collect three to five serial sections onto positively charged microscope slides using a floaton water bath set at 42°C. Dry slides overnight at 37°C on a slide warmer. Store slides desiccated at 4°C until use.

Examining Placentation in the Mouse

189

11. Place slides in rack and deparaffinize sections by immersing slides in a slide-staining dish containing histoclear for 2 × 10 min. 12. Rehydrate sections by immersing slides for 2 min in each of the following: 100% ethanol; 100% ethanol; 95% ethanol in H2O; 85% ethanol in H2O; 70% ethanol in H2O; 50% ethanol in H2O; 30% ethanol in H2O and 100% H2O. 13. Rehydrated sections should be used immediately for the analysis of histopathology by Hematoxylin and Eosin staining (see below and Fig. 2), or Hoechst staining of nuclei (see below and Fig. 3B). In addition, sections processed in this manner can be used for analysis of gene expression by in situ hybridization (8) or immunohistochemistry (9). 14. Hematoxylin and Eosin staining: We use a kit from Shandon Inc. to rapidly stain sections following manufactures instructions. However, Hematoxylin and Eosin can be purchased and solutions made according to standard protocols (10). 15. After staining with hematoxylin and eosin, dehydrate sections by immersing in the following for 2 min each: 95% ethanol; 100% ethanol; 100% ethanol. 16. Immerse slides in histoclear or xylenes for 2 × 5 min and then cover slip using paramount, taking care not to allow the formation of air bubbles under the cover glass. 17. Allow the slides to dry in a flat horizontal position until the paramount dries (approx 2 d). 18. View and photograph stained gestation sites using bright-field microscopy. This will allow for the assessment of general placental structure and cell morphology (see Fig. 2).

3.3. Using Hoechst Staining to Monitor Giant Trophoblast Cells 1. Place slides in rack and deparaffinize sections by immersing slides in a slide-staining dish containing histoclear for 2 × 10 min. 2. Rehydrate sections by immersing slides for 2 min in each of the following: 100% ethanol; 100% ethanol; 95% ethanol in H2O; 85% ethanol in H2O; 70% ethanol in H2O; 50% ethanol in H2O; 30% ethanol in H2O and 100% H2O. 3. Immerse slides in 2 µg/mL Hoechst dye in H2O for 2 min. 4. Immerse slides for 2 min in H2O and then an additional 2 min in running H 2O (see Note 11). 5. Blot slides dry using tissue paper and lay slides flat and allow to dry thoroughly by placing at 37°C for 1 h. 6. Once slides are completely dry, overlay sections with Canada balsam (approx 90 µL per slide), and cover slip, taking care not to allow the formation of air bubbles under the cover glass. 7. Blot away excess Canada balsam and keep slides in a flat position until completely dry (approx 2 wk). 8. View and photograph stained nuclei using ultraviolet epifluorescence. Nuclei of giant cells in particular, but diploid nuclei as well, will appear brilliant blue, allowing for easy identification and analysis of these cells (see Fig. 3B).

4. Notes 1. Four percent PFA is a commonly used fixative for the analysis of embryonic and extraembryonic tissue. Proper fixation of gestation sites with 4% PFA will preserve histological detail important for histopathological examination of the embryo and placenta. In addition, this fixation is suitable for the analysis of gene expression by in situ hybridization (8,11) and immunohistochemistry (9,12). 2. To ensure that males and females are of adequately mature reproductive status, use mice at least 7 wk of age. It is preferable to use stud males of proven reproductive competency. Reproductive productivity of males is optimized by mating each male only once or twice

190

Blackburn

Fig. 2. (See color plate 5 appearing after p. 262.) Developing placenta in the mouse. Gestation sites were collected and processed for hematoxylin and eosin staining as described in Subheading 3. (A) 7.5 dpc gestation site. At this stage, the embryo (e) is undergoing gastrulation. A mass of trophoblast stem cells derived from the trophectoderm form the ectoplacental cone (epc), which is located at the mesometrial pole of the implantation chamber. Giant trophoblast cells (gc) derived from the trophectoderm and epc surround the implantation chamber, as do the yolk sacs (ys). A bud of extraembryonic mesoderm known as the allatoic bud (a) will grow toward the chorion (ch) which will fuse with the extraembryonic ectoderm (ee). am, amnion; d, deciduum. Scale bar = 75 µm. (B) A 9.5-dpc gestation site. The fusion of the ch and ee with the a will form the labyrinthine zone (lz) of the mature chorioallantoic placenta. It is in this region where fetal and maternal circulatory systems are

Examining Placentation in the Mouse

3.

4.

5.

6. 7.

8. 9.

10.

11.

191

each week. Female reproductivity can be increased and more adequately timed using hormone superovulation protocols (13). A vaginal plug appears as a white or yellow hard and waxy deposit in the vaginal opening. Females should be examined for the presence of vaginal plugs in the morning, because the plug will dislodge with time, causing a pregnancy to be missed. The detection of a vaginal plug the morning following mating is designated 0.5 dpc on the assumption that coitus occurred around the middle of the dark cycle or 24:00. It is not uncommon for this estimate to be off by as much as 0.5 d between litters. If precise staging is required, gestation sites can be staged using established staging features of the embryo and uterus (14,15). The pregnant uterus can be separated into two halves. The half of the uterus in proximity to the mesentery that attaches the uterus to the dorsal wall of the abdomen is referred to as the mesometrium and contains the mesometrial deciduum. The half of the uterus opposing the mesometrium contains the developing embryo and is referred to as the antimesometrium and contains the antimesometrial deciduum (see Fig. 1C). A gestation site is defined as the developing embryo and its extraembryonic membranes contained within the uterus. The number of gestation sites per vial for fixation and subsequent dehydration, clearing, and infiltration will vary depending on the developmental stage or size of the gestation site. On 4.5–8.5 dpc, up to 10 gestation sites can be placed per vial, which holds 20 mL fixative or subsequent solutions. On 9.5–11.5 dpc, six gestation sites per vial should be used, on 12.5–15.5 dpc, four gestation sites per vial should be used, and on 16.5–19.5 dpc, two to three gestation sites per vials should be used. This will allow for proper fixation with 4% PFA as well as proper dehydration, clearing, and infiltration using the times described in the methods. Make 1X PBS and keep at 4°C for use in washing and mixing with ethanol for dehydration. Clearing of gestation sites is a critical and empirical step in the procedure. Overclearing will make the tissue brittle and difficult to section, whereas underclearing will cause improper infiltration of paraffin and, subsequently, pour sectioning and morphology of tissue. The times given in the protocol are optimized for gestation sites on 4.5–9.5 dpc. When analyzing older gestation sites, it may be necessary to increase the incubation times. A good rule of thumb is to increase times by 5 min for each additional dpc. Orienting the gestation site in this manner is important for obtaining sections through the implantation chamber that show the development of all the extraembryonic membranes, specialized structures of the developing placenta, and the embryo itself (Fig. 2). Washing or destaining of nuclei is critical for adequate visualization. The times described work best on gestation sites cut at 5 µm. Thicker sections may require more extensive washing. Inadequate washing results in high-background and nonspecific cytoplasmic staining, whereas overwashing will make visualization of nuclei difficult. The intensity of staining can be checked by viewing the freshly rinsed slide under the microscope at low

brought into close apposition. Distal to the lz is the junctional zone (jz) that consists of hormone-producing diploid trophoblast cells derived from the epc. A thin layer of gc surrounds the developing placenta and implantation chamber. Scale bar = 150 µm. (C) A 13.5-dpc placenta. After 9.5 dpc, there is a large growth and expansion of the lz and jz. By 13.5 dpc, the jz may appear acellular due to large amounts of glycogen accumulation in certain trophoblast cells. A line of gc is found surrounding the placenta. mc, maternal circulation; fc, fetal circulation. Scale bar = 300 µm.

192

Blackburn

Fig. 3. (See color plate 6 appearing after p. 262.) Visualization of abnormal placentation and giant trophoblast cells. (A) Hematoxylin and Eosin stained 9.5-dpc gestation site that has been genetically engineered to lack the purine catabolic enzyme adenosine deaminase (12). The absence of this enzyme causes the buildup of substrates that are embryotoxic. Therefore, the allantois does not make contact with the extraembryonic ectoderm (ee) and the placenta does not develop (compare with Fig. 2B). epc, ectoplacental cone; gc, giant trophoblast cells. Scale bar = 100 µm. (B) Wild-type 9.5-dpc gestation site stained with Hoechst according to methods. The enlarged nuclei of gc are easily detected surrounding the implantation chamber. Nuclei of diploid cells appear much smaller. ys, yolk sacs. Scale bar = 50 µm. power using ultraviolet epifluorescence. Sections can then be washed again or restained accordingly, before drying and mounting.

References 1. Soraes, M. J. Faria, T. N., Hamlin, G. P., Lu, X. J., and Deb, S. (1993) Trophoblast cell differentiation: expression of the placental prolactin family, in Trophoblast Cells: Pathways For Maternal–Embryonic Communication (Soares, M. J., Handwerger, S., and Talamantes, F., eds.), Springer-Verlag, New York, pp. 45–67. 2. Cross, J. C., Werb, Z., and Fisher, S. J. (1994) Implantation and the placenta: Key pieces of the development puzzle. Science 266, 1508–1518. 3. Peel, S. and Bulmer, D. (1977) Proliferation and differentiation of trophoblast in the establishment of the rat chorioallantoic placenta. J. Anat. 124, 675–687. 4. Guillemot, F., Nagy, A., Auerbach, A., Rossant, J., and Joyner, A. L. (1994) Essential role of Mash-2 in extraembryonic development. Nature 371, 333–336. 5. Gurtner, G. C., Davis, V., Li, H., McCoy, M. J., Sharpe, A., and Cybulsky, M. I. (1995) Targeted disruption of the murine VCAM1 gene: essential role of VCAM-1 in chorioallantoic fusion and placentation. Genes Dev. 9, 1–14. 6. Uehara, Y., Minowa, O., Morl, C., Shiota, K., Kuno, J., Noda, T., and Kitamura, N. (1995) Placenta defect and embryonic lethality in mice lacking hepatocyte growth factor/scatter factor. Nature 373, 702–705. 7. Naeye, R. L. (1992) Disorders of the placenta and decidua, in Disorders of the Placenta, Fetus, and Neonate: Diagnosis and Clinical Significance (Gay, S., ed.), Mosby Year Book, St. Louis, MD. 8. Albrecht U., Eichele, G., Helms, J. A., and Lu, H.-C. (1997) Visualization of gene expression patterns by in situ hybridization, in Molecular and Cellular Methods in Developmental Toxicology (Daston, G. P., ed.), CRC, Boca Raton, FL, pp. 22–47.

Examining Placentation in the Mouse

193

9. Johnstone, A. and Thorpe, R. (1996) Immunochemistry in Practice, 3rd ed., Blackwell Science, Oxford, UK. 10. Humason, G. L. (1979) Animal Tissue Techniques, 4th ed., W. H. Freeman, San Francisco, CA. 11. Shi, D., Winston, J. H., Blackburn, M. R., Datta, S. K., Hanten, G., and Kellems, R. E. (1997) Diverse genetic regulatory motifs required for murine adenosine deaminase gene expression in the placenta. J. Biol. Chem. 272, 2334–2341. 12. Blackburn, M. R., Knudsen, T. B., and Kellems, R. E. (1997) Genetically engineered mice suggest that adenosine deaminase is essential for early postimplantation development in the mouse. Development 124, 3089–3097. 13. Hogan, B., Beddington, R., Costantini, F., and Lacy, E. (1994) Manipulating the Mouse Embryo: A Laboratory Manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 14. Theiler, K. (1972) The House Mouse. Springer-Verlag, New York. 15. Kaufman, M. H. (1992) The Atlas of Mouse Development. Academic, London.

Palatal Dysmorphogenesis: Palate Organ Culture

195

21 Palatal Dysmorphogenesis Palate Organ Culture Barbara D. Abbott 1. Introduction Clefts of the secondary palate are among the most frequent birth defects in live-born human infants. Across the United States between 1981 and 1995, a compilation of data from states with birth defects monitoring programs shows the incidence of cleft palate without cleft lip to range from 2.01 to 14.2 per 10,000 live births (1). Thus, the formation of the secondary palate and the mechanism(s) for induction of cleft palate have been the focus of extensive research. Palatogenesis also presents an interesting model for many of the processes involved in morphogenesis in the embryo. The formation of the secondary palate requires neural crest cell migration, interaction of the palatal cells with the surrounding extracellular matrix, interaction and signaling between epithelial and mesenchymal cells, adhesion and fusion of morphological structures, which also involves cell death and transformation of cells from epithelial to mesenchymal phenotypes, and, finally, differentiation into bone and stratified oral and ciliated nasal epithelia (2). In order to study these processes and the potential of exogenous agents to disrupt them, in vitro models have been developed, including mesenchymal cell culture, epithelial cell culture, and palatal organ culture. Palatal organ culture can be a system in which the palatal shelves are supported on a membrane above the medium; our laboratory has used this method in the past (3–5). However, the model described in this chapter is a submerged culture of the entire midfacial region. This model offers several advantages over the earlier system, as it permits the palatal shelves to grow, elevate, and fuse in the culture medium. This model has been tested in our laboratory for several strains of rat and mouse, and it has supported development of the palate in human embryonic midfacial tissue (6). Examples of the application of this culture model include studies of the effects of methanol and 5-fluorouracil on palatogenesis (7–9). The information in this document has been funded wholly (or in part) by the U.S. Environmental Protection Agency. It has been reviewed by the National Health and Environmental Effects Research Laboratory and approved for publication. Approval does not signify that the contents reflect the views of the Agency, nor does mention of trade names or commercial products constitute endorsement or recommendation for use. From: Methods in Molecular Biology, Vol. 136: Developmental Biology Protocols, Vol. II Edited by: R. S. Tuan and C. W. Lo © Humana Press Inc., Totowa, NJ

195

196

Abbott

2. Materials 2.1. Preparation of Medium 1. 2. 3. 4. 5. 6. 7.

Custom Modified Biggers BGJ formulation, dry powder (Irvine Scientific, Santa Ana, CA). Sodium bicarbonate (cell culture tested). Monobasic and dibasic sodium phosphate. Sodium hydroxide solution 1 N. L-Glutamine (cell culture tested). Bovine serum albumin (BSA) Fraction V powder 98–99% albumin. Penicillin–streptomycin solution (cell culture tested) (10K units penicillin + 10 mg streptomycin/mL). 8. Bottle filter systems for filtration of medium, low-protein-binding filters (nylon or cellulose acetate membranes) 0.2 µm pore size, prefilters are helpful for large-volume applications. Sterile plastic flasks of 500 mL capacity for storage of phosphate-buffered saline (PBS) and medium. Optional 9. Fetal bovine serum (FBS). 10. Sodium selenite (cell culture tested). 11. apo-Transferrin (cell culture tested). 12. Sodium ascorbate (L-ascorbic acid). 13. All-trans-retinoic acid (cell culture tested).

2.2. Prepare Culture Flasks and Gas Medium 1. Dimethylsulfoxide (DMSO). 2. Tissue culture flasks canted neck 25 cm2 (or 75 cm2 for rat) plug seal 70 mL capacity. Note: Caps must form gas-tight seal, not the style that is ventilated by a filter cap (e.g. Corning Glassworks, Corning, NY #2510025). 3. Medical gas mixture of 50% O2, 5% CO2, 45% N2, analysis to confirm mixture purity. 4. Sterile disposable transfer pipets, individually wrapped, used to gas flasks. 5. Syringe filters, 0.2-µm pore size.

2.3. Culture Setup 1. Dissecting stereomicroscope, fiber-optic illuminator, black rubber stopper cut horizontally to provide 1⁄2 in.-high dissection pad or other suitable dark rubber pad (sterilize by storing in 70% ethanol), rocker platform with adjustable tilt rate (side-to-side motion) to fit in incubator, culture incubator to maintain 37°C temperature, vacuum system (pump) to filter medium. 2. Gauze sponges, 70% ethanol, plastic disposable Petri dishes, polypropylene tubes of various sizes, test tube rack. 3. Dissecting tools: For removal of uterus and embryos from uterus: dissecting forceps 4 in. serrated, curved (at least two); curved blunt tip 41⁄2-in. microdissecting scissors, and 4-in. curved sharp-tip microdissecting scissors. For dissection of embryo and removal of midfacial tissue: Dumont forceps #5 stainless steel (at least two pairs, have extra on hand and be aware that these points are easily damaged even in routine use); spring-loaded iridectomy scissors with very fine, small scissor blades such as the Vannas ultramicro scissors 3 in. straight (for mouse) (Roboz Surgical Instrument Co., Rockville, MD, RS-5610, or equivalent) and Vannas extremely delicate 3 in. straight sharp scissors (for rat) (Roboz Surgical RS-5883, or equivalent); disposable #11 scalpels (fine-pointed blade gives good visibility during dissection cuts); microdissecting spatula 5 in. (such as the Roboz RS-6150). All instruments are sterilized by soaking in 70% ethanol just prior

Palatal Dysmorphogenesis: Palate Organ Culture

197

Table 1 Custom Modified Biggers BGJ Formulation (with L-Glutamine 150 mg/mL, without NaHCO3) Irvine Scientific #98337 mg/L Sodium chloride Potassium chloride Glucose (dextrose) anhydrous Calcium lactate.5H2O L-Arginine HCl L-Cysteine HCl.H2O L-Glutamine L-Histidine HCl.H2O L-Isoleucine L-Leucine L-Lysine HCl L-Methionine L-Phenylalanine L-Threonine L-Tryptophan L-Tyrosine 2Na.2H2O L-Valine

6400 425 4400 600 70 135 150 150 23 40 225 40 40 60 30 43.25 50

mg/L p-Aminobenzoic acid Ascorbic Acid Nicotinic Acid Amide Phenol red, Na salt Thiamine HCl α-Tocopherol Phosphate, 2Na Potassium phosphate, monobasic KH2PO4 Magnesium sulfate, anhydrous MgSO4 Choline chloride Folic acid Inositol Riboflavin Vitamin B-12 d-Biotin Pantothenic acid, Ca salt Pyridoxal HCl

1.5 150 15 17.04 3 0.75 113 161.17 38 0.15 0.15 0.15 0.03 0.15 0.15 0.15

Source: ref. 10.

to use. Small and/or delicate instruments can be wrapped in sterile gauze and soaked with the ethanol.

3. Methods 3.1. Preparation of the Medium and PBS 1. Reconstitute Modified Bigger’s BGJ medium, formulation in Table 1 (10), with distilled, deionized water as per suppliers instructions to be 1X. (Add 14 g sodium bicarbonate and bring up to 5 L with water, check pH and osmolarity, should be pH 7.0–7.5, osmolarity 265–300 mOs/kg). Store liquid medium at 4°C up to 6 mo. 2. Phosphate-buffered saline: 35.04 g NaCl, 1.38 g NaH2PO · H2O (monobasic), 8.04 g Na2HPO4 · 7H2O (Dibasic), add distilled H2O up to 4 L, adjust pH to 7.4 with 1 N NaOH, filter sterilize with 0.2-µm filter, store at 4°C. 3. Supplement the Bigger’s BGJ medium prior to use (see Note 1 for preparation of stock solutions): 0.15 mg/mL L-glutamine 5 µL/mL penicillin/streptomycin (final concentration = 50 µg streptomycin/mL + 50 units penicillin/mL) 6 mg/mL bovine serum albumin (add last and mix by gentle inversion) Filter sterilize supplemented medium, store at 4°C and discard after 2 wk. (See Note 2 for additional supplements and Note 3 for cautions regarding supplements.)

3.2. Prepare Culture Flasks and Gas the Medium 1. Add solvent control and compound treatments to aliquots of the supplemented medium (e.g., compounds dissolved in DMSO added to give 1% DMSO final in medium; ethanol

198

Abbott

and methanol have teratogenic potential in this culture model (8), and if these solvents are used, controls with solvent in the medium are required to select a no-effect level). 2. Add 9 mL of control or treated medium to each sterile 25-cm2 culture flask (see Note 4 for larger flasks for rat palate). 3. Direct a stream of medical gas mixture (50% O2, 5% CO2, 45% N2) through a 0.2-µm pore filter (see Note 5) and flush air from the flask, replacing it with the gas mixture. If the gas flow rate is sufficient to agitate the medium, 10–15 s should be long enough to flush out the air. Immediately cap the flask tightly.

3.3. Culture Setup 1. Place flasks horizontally on a platform (elevating the neck of the flask makes it easier to put tissues inside, a test tube rack works well as a shelf). Loosen cap but leave it resting over flask opening. 2. Dissect tissues as described below, pick up tissue on spatula and insert into flask, touch to medium, and float tissue off the spatula. Up to four tissues per flask (see Note 4). 3. Gas flasks: After all flasks are loaded with tissues repeat the gassing step and cap tightly. 4. Incubate flasks at 37°C on a rocker platform (side-to-side rock rate approximately 15 tilts/min).

3.4. Culture Maintenance Replace medium and gas flasks every 24 h. In most cases, the treatments are also included with the fresh medium. 1. 2. 3. 4. 5.

Prewarm the complete medium in 37°C water bath. Remove medium from flask by pouring or suction (keep tissues inside flask). Replace with 9 mL fresh prewarmed medium. Gas as before. Replace on rocker in incubator. Cultures can be maintained 4–5 d or until palatal fusion occurs.

3.5. Dissection of the Embryonic Craniofacial Region 1. Timed pregnant female mice are killed on GD12 by CO2 asphyxiation and/or cervical dislocation (see Note 6). Embryos are removed from the uterus as quickly as possible and placed in chilled PBS in Petri dishes on ice. If possible, keep the placenta and yolk sac attached to embryo and the circulation intact. This will help delay degeneration of the tissues. 2. Dissect mid-craniofacial region (includes maxillary arch, primary palate, and secondary palatal shelves) (refer to Notes 7 and 8): a. Place the embryo on a black rubber platform (large rubber stopper that has been cut in half to reduce height to about 1⁄2 in.) and remove yolk sac, amnion, placenta with Dumont forceps. b. Use Vannas ultramicroscissors and insert one point of the scissors into the oral cavity, and with the other across the outside of the face, cut to separate the mandible and neck from the upper head (line A in Fig. 1A). Forceps at the back of the head can be used to brace or hold head in position during the cut. Make a second cut at the level of the eye to remove the upper brain tissues (line B). c. Push the other tissues aside and turn the midface over, exposing the palatal shelves (Fig. 1B). Use scalpel to remove hindbrain tissue with a cut at the level of the posterior of the palatal shelves (line C). d. If any mandible, tongue, or hindbrain remains, trim off. Eyes will remain on the final midfacial explant that is now ready to transfer to the culture flask. 3. Transfer explant to flask. Pick up the explant by sliding a dissection spatula under the explant; press down on the rubber mat to assist in pickup. Open cap of flask and insert

Palatal Dysmorphogenesis: Palate Organ Culture

199

Fig. 1. The gestation day 12 mouse head (plug day = 0) is shown before dissection for palatal organ culture. (A) The first incision is made across line A through the oral cavity to separate mandible from the head. The second incision is through line B to remove the upper cranial tissue. (B) After turning the midface over, exposing the palatal shelves a final cut is made through line C to remove remaining hindbrain tissues. The resulting midfacial region is placed in culture. spatula to touch the medium and float tissue off. Embryos from each litter can be distributed across the treatment groups/flasks by placing several flasks on the platform to receive tissues in sequence as the dissections occur. Put up to four tissues per flask.

3.6. Tissue Evaluation At intervals during culture and at the end of culture, the explants can be examined under a dissecting microscope to track the progress of palatal elevation and fusion. Palates that are adhered cannot be separated even by gentle pulling with Dumont forceps on the primary palatal edges. A palatal seam will remain intact and the adjacent tissue will tear if sufficient force is used. 4. Notes 1. Stock solutions of medium supplements: (Filter sterilize supplements a–d before aliquoting and store as indicated. These stock solutions should be good for 1 yr or more.) a. Sodium selenite: 5 mg in 5 mL H2O; dilute 1:100 to give 10 µg/mL stock and store as 1-mL aliquots at –20°C. b. apo-Transferrin: 20 mg in 2 mL H2O; aliquot 200 µL/tube and store 10-mg/mL stock at –20°C. c. Sodium L-ascorbate: 200 mg in 20 mL H2O; aliquot 1 mL/tube and store 10-mg/mL stock at –20°C. d. FBS: dilute 1:1 with medium prepared as above, aliquot 3–5 mL/tube and store 50% serum at –20°C. e. Penicillin–streptomycin: Aliquot 1–5 mL/tube of the 10000 U penicillin G/mL + 10 mg streptomycin/mL and store stock at –20°C. f. All-trans-retinoic acid (RA): Dissolve in DMSO at 1 × 10–2M stock, protect from light, store in amber vials at -80°C. Dilute stock with DMSO in serial dilutions for treatments or as a supplement. Add 1 µL RA/mL medium (DMSO=0.1%). 2. Additional supplements can include: a. 5 µL/mL Sodium ascorbate solution (50 µg/mL final) b. 1% Fetal bovine serum (50% stock 20 µL/mL medium) c. 1 µL/mL selenium solution (final concentration=10 ng/mL)

200

3.

4.

5.

6.

7.

8.

Abbott d. 1 µL/mL transferrin (final concentration=10 µg/mL) e. Glutamine can be increased to 4X normal level f. All-trans-retinoic acid (1 × 10–11M final in medium) Palatal shelves show normal development and fuse in the standard control medium without addition of the above-listed supplements. However, depending on the test compound, additional supplements may be required to provide responses to a toxicant that are similar to those observed after an in utero exposure. This seems to depend on the mechanism of action of the compound, as the secondary palatal shelves elevate and fuse in the absence of these additives. Caution should be exercised with supplements to assure that normal palatal processes occur in the presence of the additives. Excess fetal bovine serum (>1%) can interfere with palatal fusion in the culture model. Be aware that each new lot of serum, even if it is from the same supplier, needs to be tested in the model to confirm that it supports normal development. For larger species such as rat use 50 mL of medium in a 75-cm2 flask and place up to 10 tissues/flask. Increasing the number of tissues per flask will decrease the number of specimens with palatal fusion. For gassing the flasks, a syringe filter can be attached to the end of the gas tubing, then cut off the bulb end of a plastic disposable transfer pipet and attach to the outlet of the filter. Inserting the pipet halfway into the flask works well. Replace transfer pipet frequently (e.g., between sets of flasks containing different treatments). Mice can be mated overnight and checked for sperm plug the next morning which is designated GD0. The palatal shelves appear as outgrowths of the maxillary arch between GD11 and GD11.5. The palatal shelves should be clearly visible on GD12 (Theiler stage 20 or Carnegie stage 17) and should be vertical in orientation. A similar developmental stage for rats is found on GD14. Be aware that different strains of mice (and even the same strain in different breeding colonies) may differ by as much as 6–12 h in developmental staging. Confirm the best time for collecting tissues from your animals. The embryos need to be removed from the uterus as quickly as possible and will be easier to dissect if the membranes and placenta remain attached. Any embryos that become separated should be dissected first. The Petri dish is kept on ice to retain integrity of tissues as long as possible. With experience, the average time to dissect a litter can be less than 15 min. In our laboratory, two experienced operators can prepare 80–90 explants for culture in 1 h. The organ culture outcome may depend on expeditious preparation of the explants. It is not recommended to use any embryos that are on ice longer than 40–45 min, waiting to be dissected. Special note for rat palatal organ culture: Rat strains differ in their requirements for successful fusion of the palates in culture. The Fisher 344 rat performs well under the same conditions as the mice (four explants in 9 mL of medium for 4–5 d; culture starts Monday and ends on Thursday or Friday). Sprague-Dawley and CR/Lewis required the larger flasks with 50 mL of medium/10 explants and required media supplemented with 1 × 10–11M all trans-retinoic acid. The CR/Lewis rat palates also had a higher percentage of explants fuse if the gas was changed on the fourth day to 95% O2, 5% CO2, and the flasks were flushed with the gas in the morning and late afternoon. Tissues were scored the next morning (5) and 80% of the palates were fused.

References 1. Birth defects surveillance data from selected states. (1997) In Congenital Malformations Surveillance Report, A Report from the National Birth Defects Prevention Network. Teratology 56, 1–175.

Palatal Dysmorphogenesis: Palate Organ Culture

201

2. Zimmerman, E. F. (1997) Palate, in Handbook of Experimental Pharmacology 124/I: Drug Toxicity in Embryonic Development I (Kavlock, R. J. and Daston, G. P., eds.), Springer-Verlag, New York, pp. 183–208. 3. Abbott, B. D. and Pratt, R. M. (1987) Retinoids and epidermal growth factor alter embryonic mouse palatal epithelial and mesenchymal cell differentiation in organ culture. J. Craniofac. Gen. Dev. Biol. 7, 219–240. 4. Abbott, B. D. and Pratt, R. M. (1987) Human embryonic palatal epithelial differentiation is altered by retinoic acid and Epidermal Growth Factor in organ culture. J. Craniofac. Gen. Dev. Biol. 7, 241–265. 5. Abbott, B. D. and Birnbaum, L. S. (1990) Rat embryonic palatal shelves respond to TCDD in organ culture. Toxicol. Appl. Pharmacol. 103, 441–451. 6. Abbott, B. D. and Buckalew, A. R. (1992) Embryonic palatal responses to teratogens in serum-free organ culture. Teratology 45, 369–382. 7. Abbott, B. D., Lau, C., Buckalew, A. R., Logsdon, T. R., Setzer, W., Zucker, R. M, Elstein, K. H., and Kavlock, R. J. (1993) Effects of 5-fluorouracil on embryonic rat palate in vitro: Fusion in the absence of proliferation. Teratology 47, 541–554. 8. Abbott, B. D., Logsdon, T. R., and Wilke, T. S. (1994) Effects of methanol of embryonic palate in serum-free organ culture. Teratology 49, 122–134. 9. Shuey, D. L., Buckalew, A. R., Wilke, T. S., Rogers, J. M. and Abbott, B. D. (1994) Early events following maternal exposure to 5-fluorouracil lead to dysmorphology in cultured embryonic tissues. Teratology 50, 379–386. 10. Shiota, K., Kosazuma, T., Klug, S., and Neubert, D. (1990) Development of the fetal mouse palate in suspension organ culture. Acta. Anat. 137, 59–64.

Palatal Dysmorphogenesis: Quantitative RT-PCR

203

22 Palatal Dysmorphogenesis Quantitative RT-PCR Gary A. Held and Barbara D. Abbott 1. Introduction The formation of the secondary palate requires coordinated expression of genes involved in processes such as cell proliferation, differentiation, and production of extracellular matrix proteins. In order to study the mechanisms through which teratogenic agents induce cleft palate, the regulation and expression of these genes needs to be examined and compared between control and treated embryos. The expression of specific mRNAs can be examined by reverse transcription polymerase–chain reaction (RT-PCR) and this chapter presents an application of RT-PCR for quantitating the level of mRNA. The sensitivity of this method allows the investigator to evaluate gene expression in individual embryos using dissected tissues from the specific regions/ anatomic structures affected by the teratogen. Even extremely small specimens can usually be used to quantitate mRNAs for several genes. Other methods, such as the Southern blot, generally require pooling of multiple embryos and may not provide information for the isolated, specific target tissue. The quantitative method described in this chapter is derived from the method published by Vanden Heuvel (1). The modifications we introduce improved the performance in our laboratory and may enhance the reliability, specificity, and sensitivity of this approach. This introduction provides an overview of the issues of experimental design that are important to the success of the method, including internal standard design, primer selection, image acquisition and analysis. A basic introduction to the RT-PCR method can be found in one of the many resource and protocol books available on this topic, including a volume edited by McPherson et al. (2). The method presented in this chapter uses an internal standard RNA sequence (IS) as a control template that is added in known quantities to the total RNA from the embryonic tissue. The IS is specifically designed for each gene under study and the The information in this document has been funded wholly (or in part) by the U.S. Environmental Protection Agency. It has been reviewed by the National Health and Environmental Effects Research Laboratory and approved for publication. Approval does not signify that the contents reflect the views of the Agency, nor does mention of trade names or commercial products constitute endorsement or recommendation for use. From: Methods in Molecular Biology, Vol. 136: Developmental Biology Protocols, Vol. II Edited by: R. S. Tuan and C. W. Lo © Humana Press Inc., Totowa, NJ

203

204

Held and Abbott

primers for the IS are the same as those used to amplify the mRNA of interest. Our approach also uses the reverse primer selected to amplify the mRNA as the primer for the reverse transcription step, rather that poly-deoxythymidine. This modification allows us to select primers that amplify a region of mRNA distant from the poly A tail of the molecule. In many cases, we found that attempting to use the poly A primer was not as consistent for reverse transcription of sequences distant from that region. The reverse primer worked well for reverse transcription and provided a good template for the following PCR step. Other quantitative RT-PCR methods may include DNA sequences as standards and these may or may not utilize the same primers as those required to amplify the target mRNA. An advantage of using RNA for an internal standard is that this allows the efficiency of the RT reaction to be included in the evaluation. Some protocols amplify a constitutively expressed gene as an internal standard to compare with the mRNA of the gene of interest. This housekeeping or second gene approach requires that different primer sequences be used for the target mRNA and the standard. Although this allows a naturally existing mRNA to be used as the standard, this approach may lead to inaccurate results if the efficiency of amplification differs for the two primer sets (housekeeping gene and mRNA target). Also, it should be clearly demonstrated that the expression of the “standard” housekeeping gene mRNA does not change depending on the treatment being studied. The major advantage to using the RNA synthetic internal standard with an identical primer sequence as the target mRNA is that this approach eliminates issues related to variations in reverse transcription and PCR efficiency that can occur when different primer pairs are used for the standard and target amplification. The internal standard used in our method is a synthetically constructed RNA that uses the same primers as the mRNA under study. A serial dilution of the IS is added to a series of reaction tubes containing a constant amount of total RNA (Fig. 1A). Thus, both the target mRNA and the internal standard RNA are reverse transcribed and then amplified in the same PCR tube. After the RT-PCR reactions are run, the mRNA and IS products are separated by gel electrophoresis and images are acquired (Fig. 1B). The IS is designed to be slightly different in size relative to the amplified mRNA. The intensity of the IS and mRNA bands on the gel are quantitated and a ratio of IS/RNA values is calculated for each lane in the gel. A regression analysis of ratio against known IS concentration provides the point at which the ratio equals 1, and at that point, the amount of mRNA is equal to the amount of IS (a known quantity). Thus, the known IS level provides the level of mRNA as a function of the amount of total RNA in the assay. The features and design of the IS are illustrated in Fig. 1C. A spacer region of pUC19 plasmid sequence is selected to give an IS size slightly different than the target mRNA PCR product. Typically, a difference of about 10–15% works well. The IS is synthesized in a PCR reaction with the pUC plasmid as template and using two composite primers. The forward primer consists of a T7 polymerase promotor, the target mRNA forward primer, and a forward primer for pUC19. The reverse primer consists of a poly dT region, the reverse target mRNA primer, and a reverse pUC19 primer. The pUC19 primers are selected to give a spacer region of the correct length and a GC content similar to the mRNA being measured. We have found that the common cloning vector pU19 is a good source of the spacer region because it is readily available in highly purified form, has both high- and low-GC regions, and we have found it easy to find

Palatal Dysmorphogenesis: Quantitative RT-PCR

205

PCR primers that give reliable, clean amplifications. Undoubtedly, many other templates would be equally suitable. The RNA internal standard used in the quantitative RT-PCR reactions is synthesized using T7 RNA polymerase in an in vitro transcription reaction. The reverse primer is used for the reverse transcriptions instead of a poly dT primer; the poly dT can be omitted from the composite reverse primer. One of the critical steps required for successful quantitative RT-PCR is the selection of good primer pairs. There are many programs that assist in the selection of primers that avoid known problems such as extremes in GC content, hairpins, and self-annealing. One program that we have had good success with is Oxford Moleculars’ MacVector. We have found the default settings to usually give primers that amplify well with few artifactual bands. After selecting several prospective primer pairs, it is useful to scan Genbank for homologies to known genes. Although two primers rarely match more than the selected gene, any such combinations should be avoided. It is useful to try several primers pairs at this point to determine how well they amplify and how free they are from spurious bands. Spurious PCR bands are a reason to reject the primers at this early stage of protocol development, as optimization frequently does not help much with this problem. It is also very desirable to use primers that do not require a hot start to avoid procedural problems when performing the quantitative RT-PCR reactions. Attention needs to be given to acquisition of the image of the PCR products on the gel in a manner that prevents compromising the quantitative nature of the data. A method readily available in most laboratories is photography, followed by scanning of the photograph. Although, this can work well, there are many possible pitfalls. A major problem is the short range of a linear response of film to light. In conjunction with this problem, if flat-bed scanners are used for scanning the resulting negatives, as is increasing common, the limited maximum density of the scanners confounds the problem. The photographic/scanner-related problems can be avoided by carefully controlling exposures to ensure that the critical portions of the image are in the linear response range of both the film and scanning equipment. Digital imaging avoids some of the problems of photographic methods due to the linearity of charge-coupled device (CCD) detectors and the speed with which captured images can be evaluated for suitability. The most commonly used digital imaging systems are based on CCD video cameras. Slow-scan CDD cameras are another possible option. Video-based systems have the advantage of ease of focus and framing of the image. Disadvantages of video systems, when compared to slow-scan cameras, are lower sensitivity and less dynamic range. Video cameras that are capable of on-chip integration solve the sensitivity issue, and the dynamic range can be partly compensated by co-adding multiple images. In our laboratory, we have had good results using a Hamamatsu (Bridgewater, NJ) Argus 20 video imaging system, which consists of a video camera capable of on-chip integration and a dedicated controller that allows multiple images to be added together, thus increasing the dynamic range and reducing the noise levels of the image before transferring the images to a computer. We use the gel plotting macro package of NIH Image to determine the relative intensity of the DNA bands. 2. Materials (See EPA disclaimer regarding sources and products, other suppliers and products are available and may produce similar results.)

206

Held and Abbott

Fig. 1. (A) The Internal Standard (IS) is added as a dilution series to reactions containing constant amounts of the total RNA. Both the target mRNA and the IS are reverse transcribed and then amplified in the same PCR tube. The mRNA and IS amplified sequences are separated on a gel. (B) The band shown on this example are the amplified IS (upper bands) and/or mRNA (lower bands). The intensity of the IS decreases with the amount (pg) added as the intensity of the mRNA band increases. The quantitation of intensity is shown as peaks in the histogram and the ratio of upper to lower peak is plotted for each level of IS (each lane). Regression analysis provides a formula for estimation of the mRNA level at ratio = 1. In this example y = –2.069e – 2 + 0.95768x, with r2 = 1.000, and when y = 1 then x = 1.066. (C) The schematic represents the design of the IS in which pUC DNA is amplified with primers that include a T7 promoter region and that promoter is used to reverse transcribe the cDNA IS to prepare the cRNA IS for use in the RT-PCR reactions.

2.1. Total RNA Preparation 1. TRI® Reagent (Molecular Research Center, Inc., Cincinnati, OH) or similar product to prepare RNA using the Chomczynski (3) method. 2. Tissue homogenizer (Brinkmann Polytron® Homogenizer with PT-7 generator for small volumes (Brinkmann Instruments, Inc., Westbury, NY). 3. Eppendorf tubes, chloroform, refrigerated microcentrifuge, isopropanol, autoclaved pipet tips, ethanol, diethylpyrocarbonate (DEPC), DEPC-treated H2O, 0.3M sodium acetate (prepared with DEPC-treated, autoclaved water), low-volume cuvettes, and spectrophotometer to read OD260.

2.2. Primer Oligos 1. Primers for PCR (Genosys Biotechnologies, Inc., The Woodlands, TX). Either synthesize and purify or order purified oligos with sequences determined by computer search programs. Primers 48 h) (see Note 7). Stop reaction: rinse 2X in 2 mg/mL glycine in PBST at RT. Refixation: 30 min in 4% paraformaldehyde in PBS at RT. Rinse: 5X 5 min in PBST at RT. Embryos to be probed should continue to step 8. Embryos (or embryo fragments) to be used for preadsorption of the anti-digoxigenin antibody should be removed for adsorption of antibody process (see step 9). Prehybridization: For each basket, preheat vial containing 450 µL of Hyb buffer. Transfer each basket to individual vial. Incubate 1–2 h in Hyb buffer at 65°C. Hybridization (see Note 8): Transfer basket to fresh vial with fresh Hyb buffer and approximately 0.5 µg of probe. Incubate at 65°C ON. Adsorption of antibody (see Note 9): Embryos ( or embryo fragments) not to be probed should be incubated with 1:400 dilution of antibody in blocking solution (e.g., 12.5 µL of antidigoxigenin antibody and 5 mL of blocking solution) in a glass vial at 4°C ON. Day 2 Transfer each basket back to the rack sitting in glass staining dish with first wash solution (prewarmed to 65°C). Wash: 10 min 37.5% formamide, 2X SSC at 65°C (prewarmed) 10 min 25% formamide, 2X SSC at 65°C (prewarmed) 10 min 12.5% formamide, 2X SSC at 65°C (prewarmed) 10 min 2X SSC at 65°C (prewarmed) 2X 30 min 0.2 SSC at 65°C (prewarmed) 5 min 0.15X SSC, 25% PBST at RT 5 min 0.10X SSC, 50% PBST at RT 5 min 0.05X SSC, 75% PBST at RT 5 min PBST at RT Block: 1 h in blocking solution at RT. Antibody incubation: 3 h with preadsorbed antibody at a 1:4000 dilution in blocking solution at RT (or ON at 4°C). Wash: 2X rinse with PBST at RT 3X 15 min PBST at RT (Note: can be left in PBST ON at 4°C, if antibody incubation has been completed.) Day 3 Continued washing: 2X 15 min PBST at RT

268

Chin

3X 5 min alkaline phosphatase buffer Transfer embryos in each basket to individual well of a 24-well microtiter plate using a wide-bore Pasteur pipet. 14. Stain: Incubate embryos in each well with about 500 µL of Alkaline phosphatase buffer that includes substrates BCIP and NBT ( 3.5 µL BCIP stock and 4.5 µL NBT stock per 1 mL of Alk Phos buffer). Remember to cover the plate with aluminum foil to reduce the lightinduced degradation of the BCIP and NBT substrates. Add fresh substrate if you monitor the staining for long periods of time under the microscope. 15. Stop reaction: Rinse 4X in PBST. 16. Refixation: 30 min in 4% paraformaldehyde in PBS.

4. Notes 1. The most critical steps in the in situ hybridization protocol are the embryo preparation, permeabilization (digestion), and riboprobe preparation. Aliquots of the paraformaldehyde solution can be stored at –20°C for up to 2 mo but should not be kept at room temperature for more than a few days. Embryos should not be left in paraformaldehyde solution at 4°C for more than 3 d before being dechorionated and transferred to methanol. “Overfixed” embryos are very difficult to permeabilize; this lowers the eventual signal-to-background ration. Dechorionation is easier to do prior to, rather than after, the methanol dehydration. 2. Although in situ hybridization can be performed in vials or in multiwell tissue culture plates, the procedure can be greatly facilitated by using staining dishes inside which one can lower a rack capable of holding numerous mesh-bottom baskets (Fig. 4). Up to 50 embryos can be placed in each basket. Fluid exchanges are simply done by using a hemostat to lift the entire rack out of a staining dish containing the first solution and quickly insert it into a staining dish with the next solution. This maneuver is less traumatic to the embryos than manually removing solution from individual wells (e.g., with a Pasteur pipet). In addition, solution changes are more complete and uniform. 3. The permeabilization step using proteinase K often needs to customized to the particular tissue, structure, or organ of interest. For example, accessibility of the probe is especially important when investigating a deeply placed organ like the heart and of less importance when the region of interest is superficial, like the caudal fin. To vary the digestion, one can change either the proteinase K concentration or the duration of the digestion. 20 min 10 µg/mL proteinase K in PBST (embryos 24–48 h) 10 min 10 µg/mL proteinase K in PBST (embryos 12–24 h) 5 min 10 µg/mL proteinase K in PBST (embryos 6–12 h) 5 min 2.5 µg/mL proteinase K in PBST (embryos 1 Mb) have been introduced intact. One of the potential disadvantages of shperoplast fusion is the transfer of yeast genomic DNA in a subset of fused ES cells and the effect this may have on the ES cells and the YAC transgenic mice. However, thus far, cointegrated yeast genomic sequences have not affected the ability of ES cells to differentiate properly or to transmit the YAC through the mouse germline and have also not affected apparent gene function (17– 20). Another disadvantage of spheroplast fusion is the relatively time-consuming transmission of genetically altered ES cells through the mouse germline.

2.2. Materials 1. Materials for yeast growth and phenotype testing are described elsewhere (22). 2. Materials for preparation of yeast spheroplasts: a. STC: 1 M sorbitol, 10 mM Tris-HCl pH 7.5, 10 mM CaCl2; composition per 100 mL, 50 mL of 2 M sorbitol, 1 mL of 1 M Tris-HCl, pH 7.5, 1 mL of 1 M CaCl2. Bring to final volume with distilled water and autoclave. b. SPEM: 1 M sorbitol, 10 mM sodium phosphate, pH 7.5, 10 mM EDTA, pH 8.0, 30 mM β-mercaptoethanol (BME); composition per 10 mL, 5 mL of 2 M sorbitol, 0.1 mL of 1 M sodium phosphate, pH 7.5 (make by mixing sterile 1 M dibasic and monobasic sodium phosphate to the appropriate pH), 0.2 mL of 0.5 M EDTA, pH 8.0, 21 µL of 14 M BME (Sigma, St. Louis, MO, cat. no. M7522). The sorbitol, sodium phosphate, and EDTA

Production of Transgenic Mice with YACs

437

437

Fig. 1. (See color plate 2 appearing after p. 262.) Schematic representation of the three methods for the generation of YAC transgenic mice. (A) Spheroplast fusion relies on the introduction of YACs containing a mammalian selectable cassette (NEO) into ES cells by enzymatic digestion of the yeast cell wall and fusion of the resulting spheroplasts with ES cells. Please note that spheroplast fusion has typically utilized HPRT for positive selection in ES cells, but neor (NEO) can also be utilized and has been included in the figure for the sake of clarity. (B) Lipofection involves the isolation of YACs containing mammalian selectable cassettes away from the other yeast chromosomes and subsequent introduction of purified YAC DNA into mouse ES cells via lipid-mediated transfection. ES cells containing YACs are subsequently introduced into host blastocyst-staged embryos and germline transmission of the YAC through the resulting chimeras generates YAC transgenic mice. (C) Microinjection involves the isolation of YACs away from the other yeast chromosomes, purification, and concentration of the YAC DNA with modifications to limit the shearing of high-molecular-weight DNA in solution and direct transfer of YAC DNA into the male pronucleus of mouse eggs.

438

Jakobovits, Lamb, and Peterson

(SPE) are mixed together and brought to 100 mL with distilled water and autoclaved. The BME is added just prior to use. c. Zymolyase 20T: 100 mg/mL in 10 mM sodium phosphate, pH 7.5; composition per 1 mL, 100 mg zymolyase 20T (ICN, Costa Mesa, CA, cat. no. 320921) dissolved in 1 mL of 10 mM sodium phosphate, pH 7.5 (see above recipe for 1 M sodium phosphate, pH 7.5). Mix well (solution always stays turbid) and aliquot. Solution is stable for several months at –20°C. 3. Materials for ES tissue culture: a. ES media: Dulbecco’s modified Eagle’s medium (DMEM), 100 U/mL penicillin, 100 µg/mL streptomycin, 2 mM L-glutamine, 15% heat inactivated ES-qualified fetal calf serum (FCS), 100 µM BME, 103 U/mL murine leukemia inhibitory factor (LIF); composition per 500 ml, 415 mL DMEM, high glucose (4500 mg/L) with sodium pyruvate (110 mg/L; JRH Biosciences, cat. no. 51444-78P), 5 mL 100 X penicillin-streptomycin solution (Gibco-BRL, Gaithersburgh, MD, cat. no. 15140-122), 5 mL 200 mM L-glutamine (JRH Biosciences, cat. no. 59202), 75 mL ES-qualified FCS (heat inactivated at 56°C for 30 min; see Note 1), 4 µL of 14 mM BME, 50 µL of 107 U/mL LIF (ESGRO™; Gibco-BRL, cat. no. 13275-029). ES media should be stored at 4°C and can be kept for several weeks. Serum-free DMEM media (SFM) is simply DMEM lacking the serum and other additions. b. Gelatin: 0.1% gelatin; composition per 500 mL, 0.5 g gelatin (Sigma, cat. no. G1890) in 500 mL picopure water. Gelatin will go into solution upon autoclaving and is stable for several months at room temperature. For the preparation of gelatin-coated 10-cm tissue-culture plates, 10 mL of 0.1% gelatin is incubated on the plate for 15–30 min at room temperature and then aspirated just prior to plating cells. c. PEG Solution: 50% polyethylene glycol 1500 (Boehringer Mannheim, Indianapolis, IN, cat. no. 783-641), 10 mM CaCl2; composition per 5 mL, one 5-mL PEG 1500 bottle, 50 µL of 1 M CaCl2. Warm solution to 37°C prior to use. d. HAT-containing ES media: 100 µM hypoxanthine, 400 nM aminopterin, 16 µM thymidine; add 10 mL 50 X HAT reconstituted supplement (Sigma, cat. no. H0262) to 500 mL ES media. HAT selection will only work on HPRT-deficient ES cell lines such as E14TG2a or E14.TG3B1 (24) when the HPRT gene is present on the YAC (13,17,19). e. G418-containing ES media: 225 µg/mL G418 (a neomycin analog; see Note 2); add 0.56 mL 200 mg/mL G418 (Geneticin; Gibco-BRL, cat. no. 11811-031; made by dissolving to 200 mg/mL active concentration G418 in 20 mM HEPES, pH 7.4 followed by filter sterilization, aliquoting and storage at –20°C) to 500 mL ES media. G418 selection will only work when the neor gene is present on the YAC.

2.3. Methods 1. Growth of yeast strains containing YACs: Inoculate a fresh colony of a YAC-containing yeast strain from selective plates into 10 mL selective media. Grow O/N at 30°C at 250 rpm on an orbital shaker (see Note 3). Transfer the growing culture to 500 ml of selective media and grow for an additional 24 h (see Note 4). Count cells with a hemocytometer to determine density (there should be approx 107 cells/mL or an OD600 of 1; see Note 5). 2. Preparation of yeast spheroplasts: Pour desired amount of culture (2.5 × 108 yeast spheroplasts are needed for each fusion) into 50-mL sterile conical falcon tubes and pellet cells in a Sorvall table-top centrifuge at 3000 rpm for 5 min. Wash cells once with 20 mL sterile water and resuspend pellet by vortexing or by triturating with a pipet. Pellet cells as described earlier and resuspend in 20 mL of 1 M sorbitol. Pellet cells once again and resuspend yeast cells in SPEM at 5 × 108 cells/mL. Take 10 mL of the sample into 90 mL

Production of Transgenic Mice with YACs

439

of 5% SDS and another 10 mL of the sample into 1 M sorbitol. Count cells from each sample—they should be the same density (see Note 6). Warm yeast suspension at 30°C. Add 1.5 µL zymolyase 20T/mL of yeast cells and gently mix (see Note 7). Incubate cells at 30°C taking 10 µL samples into 90 µL 5% sodium dodecyl sulfate (SDS) every 5 min as described previously. Determine the drop in cell number and continue until approx 90% of the cells are lysed in 5% SDS. Spheroplasts can take 5–20 min for completion (see Note 7). Immediately pellet the spheroplasted cells at 1000 rpm for 5 min. Pour off supernatant carefully—the pellet should be very loose and some loss of cells will occur. Resuspend pellet by gentle trituration with a pipet (do not vortex). Wash spheroplasts twice with 20 mL STC, gently resuspending the pellet each time. Finally, resuspend the spheroplasts in STC at 2.5 × 108 cells/mL. Keep spheroplasts at room temperature until ready to fuse to ES cells. 3. Preparation of ES cells and fusion with yeast spheroplasts: A detailed description of the steps involved in ES cell manipulations and generation of transgenic mice utilizing ES cells are beyond the scope of this chapter (for a detailed description, see refs. 25–27). Plate ES cells at 6 × 106 cells/10-cm plate, coated with a mouse primary fibroblast feeder layer. After approx 48 h, trypsinize the cultures and plate the cells onto gelatin-coated plates at 107 cells/10-cm plate for 16–24 h. Feed ES cultures 4 h before fusion. Trypsinize ES cells, wash three times with SFM (pellet at 1500 rpm for 5 min in tissue culture centrifuge) and bring to a final concentration of 5x106 cells/ml in SFM. Pellet 1 mL (2.5 × 108) of yeast spheroplasts in a 15-mL conical tube at 1500 rpm for 5 min. Aspirate all STC buffer and apply gently (horizontal as possible) 1 mL of ES cells (5 × 106 cells), without disturbing the yeast pellet (see Note 8). Pellet ES cells at 1200 rpm for 3 min and aspirate all media. The copellet can be loosened by gently tapping on tube. Add 0.5 mL of PEG solution drop-wise to the copellet while gently mixing with the 1-mL pipet tip and then pipet once gently. Incubate at room temperature for 90 s and then slowly add 5 mL SFM from the bottom of the tube and incubate for 30 min at room temperature. Spin cells at 1200 rpm for 3 min and resuspend in 10 mL complete ES media and plate 5 × 106 cells (one tube) onto one 10-cm feeder-coated plate (neor feeders, if G418 selection is planned). Change to fresh media the following morning and start drug (HAT, G418) selection 48 h postfusion. Change media every 2–3 d and pick ES colonies at 10–15 d postfusion onto feeder-coated wells for expansion and for DNA (see Note 9) and/or gene expression analysis. ES cells containing integrated, intact YACs are utilized to produce chimeric mice by standard protocols (26), and transmission of the YAC through the germline results in the production of YAC transgenic mice.

3. Lipofection

3.1. Introduction Lipofection requires the isolation of YACs from yeast cells, their purification away from the other yeast chromosomes by preparative pulsed-field gel electrophoresis (PFGE), the complexing of the YAC DNA with various lipid reagents and transfection into ES cells (see Fig. 1B) (23,28–33). The selection for the presence of the YACs in ES cells can be accomplished through the previous introduction of selectable cassettes (HPRT, neor) into the YAC vector arm via homologous recombination in yeast (protocols detailing homologous recombination in yeast as well as general yeast methods are described elsewhere; see refs. 12,22,23). Lipofection with YACs containing a selectable cassette results in approx 10–30% of the drug resistant ES colonies retaining an integrated, intact, and unrearranged copy (or multiple copies) of the YAC. In addition, selected ES clones frequently contain fragmented YACs. The overall efficiency of

440

Jakobovits, Lamb, and Peterson

lipofection varies considerably depending on the YACs, the DNA preparation and the ES cell lines utilized, but generally 1–20 drug-resistant ES colonies are obtained per microgram of input YAC DNA [(29); K. A. Bardel, L. A. Shaprio, and B. T. Lamb, unpublished observations]. YACs up to 1000 kb in length have been introduced intact into transgenic mice via lipofection (71; K. A. Bardel, L. A. Shapiro, and B. T. Lamb, unpublished observations). The lipofection of purified YACs into ES cells has its advantages; for example, no (or little) yeast DNA is transferred into ES cells, the integrity and expression of genes on the YAC can be determined in ES cells prior to introduction into mice, and large YACs (>1 Mb) have been introduced intact. The disadvantages include the often tedious isolation of YAC DNA by PFGE, the fragmentation of YAC DNA during isolation and transfection and the relatively long time for transmission of modified ES cells through the mouse germline.

3.2. Materials 1. Materials for yeast growth and phenotype testing are described elsewhere (22). a. YPD media: 1% yeast extract, 2% peptone, 2% dextrose; composition per liter: 10 g yeast extract (Difco), 20 g peptone (Difco), 20 g dextrose. Dissolve components in 800 mL distilled water, bring volume up to 1 L, and autoclave 20–40 min, 121°C, 15 psi. For solid media, add 20 g of bactoagar (Difco) prior to autoclaving. 2. Materials for agarose plug preparation: a. Solution A: 50 mM EDTA; composition per liter: 100 mL of 0.5 M EDTA, pH 8.0 (disodium salt), 900 mL distilled water. b. Solution B: 1 M Sorbitol, 20 mM EDTA, pH 8.0, 10 mM Tris-HCl, pH 7.5, 14 mM BME, 1 mg/mL zymolyase 20T: composition per 50 mL, 25 mL of 2 M sorbitol, 2 mL of 0.5 M EDTA, pH 8.0, 0.5 mL of 1 M Tris-HCl, pH 7.5, 50 µL of 14 M BME (Sigma, cat. no. M7522), 50 mg zymolyase 20T (ICN, cat. no. 320921). Bring up to final volume with sterile water. This solution should be made fresh just prior to use. c. Solution C: 1 M sorbitol, 20 mM EDTA, pH 8.0, 10 mM Tris-HCl, pH 7.5, 14 mM BME, 1% low melting point (LMP) agarose; composition per 50 mL: 25 mL of 2 M sorbitol, 2 mL of 0.5 M EDTA, pH 8.0, 0.5 mL of 1 M Tris-HCl, pH 7.5, 50 µL of 14 M BME, 0.5 g SeaPlaque GTG LMP agarose (FMC, Rockland, ME, cat. no. 50110). The sorbitol, EDTA, Tris-HCl, and LMP agarose are brought up to volume with sterile water and heated to a boil in a microwave until the agarose is completely dissolved. The BME is then added and the solution is equilibrated at 37°C. d. Solution D: 1% lithium dodecyl sulfate (LDS), 100 mM EDTA, pH 8.0, 10 mM TrisHCl, pH 8.0; composition per liter: 10 g dodecyl lithium sulfate (Sigma, cat. no. L4632), 200 mL of 0.5 M EDTA, pH 8.0, 10 mL of 1 M Tris-HCl, pH 8.0. Bring up to volume with sterile water. This solution can be stored at room temperature for several months. e. Solution E: 50 mM EDTA, pH 8.0, 10 mM Tris-HCl, pH 8.0; composition per liter: 100 mL of 0.5 M EDTA, pH 8.0, 10 mL of 1 M Tris-HCl, pH 8.0. Bring up to volume with distilled water. This solution can be stored at room temperature for several months. 3. Materials for preparative PFGE and isolation of YAC DNA: a. 5X TBE: 45 mM Tris, 45 mM boric acid, 1 mM EDTA, pH 8.0; composition per liter: 54 g Tris base, 27.5 g boric acid, 20 mL of 0.5 M EDTA, pH 8.0. Bring up to volume with distilled water. b. Agarose: SeaPlaque GTG and Nusieve GTG agarose (FMC, cat. no. 50080). c. YAC stabilization buffer: 20 mM Tris-HCl, pH 7.5, 1 mM EDTA, pH 8.0, 0.5 mM spermine; composition per 100 mL: 2 mL of 1 M Tris-HCl, pH 7.5, 200 µL of 0.5 M

Production of Transgenic Mice with YACs

441

EDTA, pH 8.0, 50 µL of 1 M spermine tetrahydrochloride (Sigma, cat. no. S2876; dissolved in water and filter sterilized). Bring up to volume with sterile water. d. β-agarase I (1 U/µL; New England Biolabs, Beverly, MA, cat. no. 392L). 4. Materials for ES tissue culture: a. ES media: DMEM, 100 U/mL penicillin, 100 µg/mL streptomycin, 0.1 mM non-essential amino acid solution, 1 mM sodium pyruvate, 15% heat inactivated ES-qualified fetal calf serum, 100 µM BME, 103 U/mL murine LIF; composition per 500 mL, 405 mL DMEM, high glucose with L-glutamine (Gibco-BRL, cat. no. 11965-092), 5 mL of 100X penicillin-streptomycin solution (Gibco-BRL, cat. no. 15140-122), 5 mL of 10 mM MEM nonessential amino acids solution (Gibco-BRL, cat. no. 11140-076), 5 mL of 100 mM MEM sodium pyruvate solution (Gibco-BRL, cat. no. 11360-070), 75 ml of ES-qualified fetal calf serum (Gibco-BRL, cat. no. 16141-079, heat inactivated at 56°C for 30 min; see Note 1), 5 mL of 0.1 mM BME (7 µL of 14 M BME stock mixed in 10 mL of picopure water, filter sterilized), 50 µL of 107 U/mL LIF (ESGRO™; Gibco-BRL, cat. no. 13275-029). ES media should be stored at 4°C and can be kept for several weeks. b. Reduced serum media: OptiMEM-I media (Gibco-BRL, cat. no. 31985-070). c. Lipid transfection reagent: Lipofectin (Gibco-BRL, cat. no. 18292-037; 1 µg/µL). d. G418-containing ES media; see Subheading 2.2., item 3.

3.3. Methods 1. Growth of yeast strains containing YACs: Inoculate a fresh colony of a YAC-containing yeast strain from a selective plate into 8 mL of selective media. Grow O/N at 30°C in a roller-drum or on an orbital shaker (225 rpm). Transfer the growing culture (the OD600 should be approx 1; see Note 5) to 200 mL of culture of YPD and grow for an additional 24 h. Count cells with a hemocytometer to determine density (there should be approx 108 cells/mL or an OD600 of 10; see Note 5). 2. Preparation of high-molecular-weight DNA plugs for preparative PFGE: Pellet cells in 250-mL bottles in a Sorvall GSA rotor at 2500 rpm for 15 min. Resuspend cells in 100 mL of solution A by pipeting and/or vortexing and pellet again. Resuspend cells in 3–4 mL of solution B warmed to 37°C (see Note 10). Add an equal volume of solution C cooled to 37°C and mix by gentle trituration with a 10-mL pipet and immediately aliquot onto strips of bent parafilm (V-shaped) sitting on ice (see Note 11). Let plugs solidify for 5–10 min and cut into 3–5-mm slices with a clean razor. Slide plugs into a 50-mL conical tube, add sufficient solution B to cover the plugs (6–8 mL), and incubate at 37°C for 2 h (see Note 12). Carefully remove buffer from plugs, add 25 mL buffer D and incubate at 37°C for 1 h. Replace with 25 mL of fresh buffer D and incubate at 37°C O/N. Wash four times with 25 mL of solution E at room temperature for 1 h on a slow-speed orbital shaker (see Note 13). Plugs may be stored at 4°C for up to 1 yr. To determine the quality of the plug DNA for YAC isolation, a standard PFGE is run (CHEFDR III [Bio-Rad, Hercules, CA, cat. no. 170-3695], 1% Seakem LE agarose [FMC], 0.5X TBE, 12°C, 6 V/cm, 120° angle, 60-s switch for 14 h, 90-s switch for 10 h). 3. Preparative PFGE and YAC DNA isolation: Assemble preparative gel by placing a 30-well comb (Bio-Rad, cat. no. 170-3628) with all of the teeth taped together in a large casting stand (21 × 14 cm; Bio-Rad, cat. no. 170-3704). Position comb so that it is in complete contact with the bottom of the casting stand. It is imperative that the casting stand is clean and level. Prepare 225 mL of 1% SeaPlaque GTG agarose in 0.5X TBE by completely melting in microwave oven and pour into casting stand after cooling slightly. After the gel has solidified at room temperature, place at 4°C for 15–20 min to completely harden. Place gel in clean, dry, level chamber, carefully remove the comb, and gently pull apart the top portion of the gel so that the well is much wider in the middle and yet is still

442

Jakobovits, Lamb, and Peterson

connected at both ends of the gel. Remove any remaining buffer in the well by gently blotting with paper towels. Load preparative gel by completely melting 3–3.5 mL of plugs at 68–70°C for 15–20 min in a 17 × 100-mm polypropylene round-bottom tube (Falcon, Lincoln Park, NJ, cat. no. 2059) and slowly add to the center of the well using a 5-mL serological pipet with a few millimeter of the tip removed with a hot razor (see Note 14). Using a wall of the casting stand as a guide, carefully lower the top edge of the well so that it is even with the sides of the gel. Let the plugs solidify in the well for 10–15 min, and then add 2 L of chilled 0.5X TBE (see Note 15). Run preparative PFGE (CHEF-DR III, 1% Seaplaque GTG agarose, 0.5X TBE, 12°C, 6 V/cm, 120° angle; conditions will vary dependent on the size of the YAC; see Note 16). Remove gel from apparatus, cut 15-mm strips from the both sides of the gel and store remainder covered with saran wrap at 4°C. Stain the gel strips in 0.5X TBE with 1 µg/mL ethidium bromide for 30 min on a rotating platform. Destain in water for 30 min at room temperature (see Note 17). Locate the YAC band under ultraviolet (UV) illumination and mark the position by slightly cutting the gel on either side of the band with a razor. Reassemble the gel and carefully cut out the region of the unstained preparative gel corresponding to the YAC band. In addition, cut out a band corresponding to a neighboring yeast chromosome to serve as a marker lane for the second-dimension gel electrophoresis. Cut excised bands into four equal-sized pieces (approx 4.5 cm) and store in solution E (see Subheading 3.2., item 2). Excised bands can be stored for several weeks at 4°C. Stain and destain the remaining gel to ensure that the YAC band was properly excised. To concentrate the YAC DNA (34) (see Note 18), place the four gel slices together in an 12 × 14 cm (Owl, Woburn, MA, model B2) gel casting stand at a 90° angle to the PFGE, along with a single gel slice of the marker yeast chromosome on either side of the YAC-containing gel slices. Prepare 150 mL of 4% Nusieve GTG agarose in 0.5X TBE, melt in the microwave oven, and cool slightly. Pour into casting stand until the molten gel is level with the top of the gel slices and let solidify. Electrophorese in 800 mL of 0.5X TBE at 6 V/cm (84 V for a 14 cm gel) for 12–14 h at 4°C. Excise the marker lanes and stain (see above) to localize and mark the position of the DNA within the Nusieve gel. Remove the corresponding location in the YAC-containing samples (0.5–1 cm) and place in solution E (see Subheading 3.2., item 2). The excised, concentrated YAC band can be stored for several weeks at 4°C. To release the YAC DNA from the agarose, first determine the weight of the agarose plug. Place the plug in a 17 × 100 mm polypropylene round-bottom tube (Falcon) and wash twice for 30 min at 4°C in 5 mL of YAC stabilization buffer (see Note 19), followed by an O/N wash at 4°C. Transfer the agarose plug to a new 15-mL tube, being careful to remove all of the buffer and melt at 68°C for 15–20 min (see Note 20). Equilibrate sample at 40°C for 15 min and add 3 U β-agarase I/100 mg of gel by gently swirling in with pipet. Prewarm the appropriate volume of β-agarase I at 40°C prior to addition. Incubate for 1 h at 40°C, gently swirl the mixture with a pipet, and continue incubation for another 1–2 h. To determine whether the agarose is completely digested, place the tube on ice for 10 min (see Note 21). To determine the concentration and integrity of the DNA, run 5–10 µL on a 0.8% standard agarose gel using λ/HindIII makers as a known concentration standard, and 30–40 µL analyzed by PFGE, respectively. YAC DNA is agarased just prior to transfection and is stored at 4°C for no more than a few days. 4. Lipid-mediated transfection of ES cells: A detailed description of the steps involved in ES cell manipulations and generation of transgenic mice utilizing ES cells are beyond the scope of this chapter (for a detailed description, see refs. 25–27). Plate ES cells at 3 × 106 cells/10-cm plate, coated with a mouse primary fibroblast feeder layer. After 48 h,

Production of Transgenic Mice with YACs

443

trypsinize ES cells, resuspend in complete ES media (without penicillin-streptomycin; see Note 22) and gently triturate to form a single-cell suspension. Count cells with a hemocytometer. Spin down ES cells at 1500 rpm for 5 min in a tissue-culture centrifuge and resuspend at 2.5 × 106 cells/mL in Optimem I media. Add 1 mL of purified YAC DNA (approx 1–2 µg) to a 17 × 100-mm polystyrene round-bottomed tube (Falcon, cat. no. 2057) using a wide-bore pipet. Mix in 50 µL of lipofectin by gently swirling the YAC DNA solution with the pipet (see Note 23). Incubate mixture at room temperature for 45 min to form lipid-DNA complexes. Aspirate media from a 10-cm tissue-culture plate coated with a primary fibroblast feeder layer and add 8 mL Optimem I media. Add 1 mL ES cell suspension (2.5 × 106 cells) to the plate and then slowly add the lipid-DNA complexes while gently swirling the plate. Place plates in 37°C incubator for 4 h. Add 10 mL complete ES media (without penicillin-streptomycin), bringing the final volume up to 20 mL. Place plates back into incubator for 14–16 h and then replace the media and lipid-DNA complexes with 10 mL of complete ES media (with penicillin-streptomycin). At 24–36 h posttransfection trypsinize thecells and place onto new neor feeder-coated 10-cm plates and add complete ES media (with penicillin streptomycin) with 225 µg/mL G418. Feed plates every 2–3 d with G418-containing ES media and pick ES colonies at 10–15 d posttransfection onto feeder-coated wells for expansion and for DNA (see Note 9) and/or gene expression analysis. ES cells containing integrated, intact YACs are utilized to produce chimeric mice by standard protocols (26), and transmission of the YAC through the germline results in the production of YAC transgenic mice.

4. Microinjection 4.1. Introduction YACs (as well as P1s and BACs) can also be introduced into mice by direct microinjection into single-cell mouse embryos (see Fig. 1C) (34–61). The preparation of YAC DNA suitable for microinjection has seen several major advances: purification by preparative PFGE, enzymatic digestion of agarose in the presence of high salt and/ or polyamines to protect against the shearing of high-molecular-weight DNA in solution and concentration (up to 1–5 ng/mL) by low-speed ultrafiltration (23,35,62), dialysis with sucrose (36), or a second-dimension electrophoresis (34,63). The concentrated YAC DNA is subsequently injected into the male pronucleus and offspring are scored for the presence of various portions of the YAC. The efficiency of introducing YACs into mice via microinjection has varied considerably between studies, but the most complete published data (34,61) suggests that although the number of transgenics per liveborn pups is similar to what is observed for standard transgenesis (approx 10–25%), a much larger number of injected oocytes (five- to 10-fold) are required to generate the same number of liveborn pups. These observations suggest that there is a certain toxicity/lethality associated with the microinjection of purified YAC DNA into mouse embryos. YAC transgenic mice produced by microinjection frequently contain fragmented YACs with single or multiple copies of DNAs integrated at either one or multiple sites in the genome. YACs up to 670 kb in length have been introduced intact into transgenic mice via microinjection (56). The microinjection of YACs into single-cell mouse embryos has a number of advantages: the production of transgenic mice via microinjection is relatively rapid, no (or little) yeast DNA is transferred into mice, and any resulting mice can be tested immediately for the complementation of existing mouse mutants (37,41,44,47) or other phe-

444

Jakobovits, Lamb, and Peterson

notypic effects. The disadvantages include the tedious and technically difficult isolation and concentration of intact YAC DNA, the shearing and fragmentation of YAC DNA during microinjection, and the possibility that larger YACs (>1 Mb) may be extremely difficult to microinject because of shearing and the physical constraints of molarity of extremely large DNA molecules.

4.2. Materials (see Note 24) 1. Materials for yeast growth: a. YPD media: see Subheading 3.2., item 1. 2. Reagents for agarose plug preparation: a. Plug molds (Bio-Rad). b. 50 mM EDTA, pH 8.0: Dilute 0.5 M EDTA, pH 8.0 stock 1:10. c. Solution 1: 1 M sorbitol , 20 mM EDTA, pH 8.0, 14 mM β-mercaptoethanol (BME), 2 mg/mL zymolyase-20T; composition per 10 mL: 10 mL of 1 M sorbitol, 400 µL of 0.5 M EDTA, pH 8.0, 10 µL of 14 M BME (Sigma, cat. no. M7522), 20 mg zymolyase-20T (ICN, cat. no. 320921). This solution should be made fresh just prior to use. d. Solution 2: 1 M sorbitol, 20 mM EDTA, pH 8.0, 2% SeaPlaque GTG low-melting point agarose (LMPA; FMC, cat. no. 50110), 14 mM BME; composition per 10 mL: 10 mL of 1 M sorbitol, 400 µL of 0.5 M EDTA, pH 8.0, 0.2 g agarose, 10 µL of 14 M BME. The sorbitol, EDTA, and LMPA are heated to a boil and equilibrated to 50°C, the BME is added and solution is held at 50°C until needed. e. Solution 3: 1 M sorbitol, 20 mM EDTA, pH 8.0, 10 mM Tris-HCl, pH 7.5, 14 mM BME, 2 mg/mL zymolyase-20T; composition per 100 mL: 100 mL of 1 M sorbitol, 4 mL of 0.5 M EDTA, 1 mL of 1 M Tris-HCl pH 7.5, 100 µL of 14 M BME, 200 mg zymolyase-20T. Make fresh day of use. f. LDS: 1% Dodecyl lithium sulfate, 100 mM EDTA, pH 8.0, 10 mM Tris-HCl, pH 8.0; composition per 500 mL: 5 g dodecyl lithium sulfate (Sigma, cat. no. L4632), 100 mL of 0.5 M EDTA, 5 mL of 1 M Tris-HCl, pH 8.0, and water to 500 mL. Filter-sterilize; can store at room temperature for several months. g. 1X NDS: 0.5 M EDTA, 10 mM Tris, 1% N-laurylsarcosine, pH 9.0; composition per 500 mL: 93 g EDTA (di-sodium salt), 0.6 g Tris base, 5 g N-laurylsarcosine. Mix EDTA and Tris base in 350 mL of water, adjust pH to greater than 8.0 with 100–200 NaOH pellets. Add N-laurylsarcosine (predissolved in 50 mL water) and pH to 9.0 with 5 M NaOH solution. Bring volume up to 500 ml with water and filter-sterilize. Store at 4°C for up to 1 yr. Working stock is 0.2X NDS; dilute to this concentration using sterile water. This solution may be stored as for 1X NDS. h. TE, pH 8.0: 10 mM Tris-HCl, pH 8.0, 1 mM EDTA, pH 8.0; composition per liter: 10 mL of 1 M Tris-HCl, pH 8.0, 2 mL of 0.5 M EDTA, pH 8.0. 3. Materials for preparative PFGE and isolation of YAC DNA: a. 0.5X TBE: See Subheading 3.2., item 3, for recipe for 5X TBE and dilute accordingly. Autoclave before use. b. 2% Absolve solution (NEN Dupont, Boston, MA, cat. no. NEF971). Prepare according to manufacturer’s instructions. c. Agarose MP (Boehringer-Mannheim, Indianapolis, IN, cat. no. 1-388-983). d. Seaplaque GTG, Nusieve GTG, and Seakem GTG agarose (FMC, cat. no. 50110, 50080, and 50070, respectively). e. Injection buffer: 10 mM Tris-HCl, pH 7.5, 250 µM EDTA, pH 8.0, 100 mM NaCl; composition per liter: 10 mL of 1 M Tris-HCl, pH 7.5, 0.5 mL of 0.5 M EDTA, pH 8.0, 20 mL of 5 M NaCl. Autoclave solution before use.

Production of Transgenic Mice with YACs

445

f. β-agarase I (1 U/µL; New England Biolabs, cat. no. 392L). g. 0.2 µM Acrodisk syringe filters (Gelman, Ann Arbor, MI, cat. no. 4192).

4.3. Methods 1. Growth of yeast strains containing YACs: Inoculate 200 mL YPD with 1 mL of an O/N culture of the yeast strain containing the YAC. Incubate O/N at 30°C with shaking (225– 250 rpm). Chill appropriate number of plug molds on ice (see Note 25). Count cells with a hemocytometer. The culture should be saturated at approx 1 × 108 cells/mL. 2. Preparation of high-molecular-weight DNA plugs for preparative PFGE: Spin down cells at 600g (see Note 26) for 5 min in a 250-mL polypropylene centrifuge bottle at room temperature. Pour off media and resuspend pellet in 80 mL of 50 mM EDTA, pH 8.0. Spin-down cells again, pour off supernatant, and resuspend pellet in 20 mL of 50 mM EDTA, pH 8.0. Transfer suspension into a preweighed 50-mL Falcon tube. Spin-down cells again (see Note 27) and decant off supernatant. Weigh the yeast pellet and tube, subtract the weight of the tube, and assume a density of 1 (weight = volume) for the yeast pellet. Add solution 1 to give 8 × 109 cells /mL. The yeast will comprise 75–90% of the volume. Add an equal volume of solution 2. The yeast concentration in the plug will be 4 × 109 cells/mL. Mix rapidly, without introducing bubbles and pipet aliquots into chilled plug molds with a blue pipet tip. Leave on ice for 10 min to allow agarose to set. Transfer plugs to 50 mL Falcon tube containing 40 mL of solution 3 (see Note 28). Incubate at 37°C for 2 h with occasional gentle rocking. Remove supernatant with 25-mL pipet, add 40 mL of LDS (see Note 28) and incubate at 37°C for 1 h with occasional gentle rocking. Replace LDS with fresh solution and incubate at 37°C O/N without agitation. Remove LDS and add 40 mL of 0.2X NDS (see Note 28). Incubate at room temperature for 2 h with agitation. Replace 0.2X NDS with fresh solution and incubate as just described. Wash plugs twice with 40 mL of TE, pH 8.0 (see Note 28) for 30 min at room temperature with agitation. Remove second TE wash and add 20 mL of fresh TE, pH 8.0. Plugs may be stored at 4°C for at least 1 yr. 3. Preparative PFGE and YAC isolation: Day 1: Prepare CHEF Gel apparatus for PFGE. Autoclave 4 L of 0.5X TBE the day before the PFGE run. Cool to room temperature or 4°C, because the temperature will be maintained at 12°C during the electrophoresis run. Set up gel system (CHEF-DR II; Bio-Rad cat. no. 170-3612). Make sure that the electrophoresis chamber is level using a bubble level. Circulate 1 L of sterile water through the electrophoresis chamber and refrigeration unit for 30 min to 1 h to clean the unit prior to adding sterile 0.5X TBE. Drain water from PFGE apparatus and fill with 2 L sterile 0.5X TBE. Equilibrate to 12°C. Cast the gel for PFGE. Soak gel casting stand parts and comb in 2% Absolve for 1 h. Rinse well with sterile water. Assemble gel casting stand wearing gloves. Tape enough teeth of the comb to obtain a preparative lane of approx 7 cm (five to six teeth of Bio-Rad 15-tooth comb; see Note 29) or use a preparative comb (Bio-Rad, cat. no. 170-3623). Casting stand should be level. Prepare 200 mL of 0.5% agarose MP solution: 1 g agarose in 200 mL of sterile 0.5X TBE. Melt by boiling on hot plate or in microwave oven. Cool to 58°C and pour. After the gel has solidified and reached room temperature, refrigerate for 15–30 min. This will prevent the wells from collapsing in this low percentage gel when the comb is removed. Load preparative yeast plugs cut to 7-mm height next to one another in the preparative well. In the standard size wells flanking the preparative well load a small amount of the preparative plug, yeast chromosome markers (New England Biolabs or BioRad), and λ midrange marker I (New England Biolabs). Seal all wells with sterile 0.8% Seaplaque LMPA in 0.5X TBE. Run the PFGE using the following conditions:

446

Jakobovits, Lamb, and Peterson

CHEF-DR II (Bio-Rad), 12°C, 200 V, 60-s switch for 20 h (see Note 30). We have used these conditions to successfully isolate 155-kb, 248-kb, 450-kb, and 650-kb YACs. Day 2: Locate the YAC band. When the gel run is complete, cut off the marker lanes flanking the preparative lane and stain them on a rotating platform in 0.5X TBE containing 0.5 µg/mL ethidium bromide for 30 min. Destain in 0.5X TBE for 30 min. Locate the YAC band under UV illumination (long or short wave) and mark the position by cutting out a notch encompassing the YAC band with a sterile scalpel, razor blade, or glass cover slip. Reassemble the gel and cut out the unstained region of the preparative lane containing the YAC DNA. Cut out a second region of the preparative lane encompassing a yeast chromosome of approximately the same size as the YAC to serve as a marker lane for the second gel run. Stain and destain the remaining gel as described above to ensure that the YAC band was properly excised. Concentration of YAC DNA by second-dimension agarose gel electrophoresis (see Note 31). Presoak a mini-gel apparatus and gel casting stand in 2% Absolve for at least 1 h or O/N on d 1. Prepare 50 mL of 4% Nusieve GTG agarose solution: 2 g agarose in 50 mL of sterile 0.5X TBE. Melt by boiling on hot plate or in microwave oven. Cool to 68–72°C before pouring. Place YAC slice and marker slice in the mini-gel casting stand at a 90° angle to the PFGE run so that the DNA will migrate the length of the PFGE gel slice into the Nusieve gel. Pour the gel around them. Electrophorese in sterile 0.5X TBE at 2.4 V/cm (48 V for 20 cm mini-gel apparatus) for 16 h (see Note 32). Day 3: Excise YAC plug. Excise the marker lane and stain as described above to localize the DNA within the Nusieve gel. Locate the corresponding position in the YACcontaining lane and cut out a Nusieve gel plug containing the YAC (usually 0.5 cm; see Note 33). Stain the remaining YAC lane to demonstrate excision of the YAC plug. Place YAC slice in a preweighed 50-mL Falcon tube and weigh to determine the weight of the plug. Add 40–50 mL injection buffer and equilibrate for 1–2 h at room temperature with occasional rocking (see Note 34). Agarase treatment: Transfer the gel slice to a sterile, DNase-free 1.5-mL microfuge tube with sterile forceps. Melt the agarose at 68°C for 10 min, rapidly transfer to 42.5°C and equilibrate for 5 min at this temperature. Add 2 U of β-agarase I/100-mg gel slice using a wide-bore yellow pipet tip. Prewarm the appropriate volume of agarase to room temperature in the yellow tip prior to addition. Mix by gently releasing air bubbles into the solution three times from a wide-bore yellow pipet tip. Incubate for 3 h (or up to O/N) at 42.5°C. Place on ice for 10 min to check for undigested agarose. Only if many lumps are present should the agarase treatment be repeated. Small amounts of undigested agarose are removed by filtration just prior to microinjection. To determine DNA concentration, check 5 µL on a 0.8% agarose gel using λ/HindIII markers of known concentration as a standard. Check the integrity of the YAC DNA by running 20–25 µL of purified DNA on a PFGE gel followed by Southern-blot analysis. The agarose and PFGE conditions should be chosen to allow optimum resolution of the YAC to be microinjected. 4. Microinjection into fertilized mouse oocytes: A detailed description of the steps involved in murine transgenesis are beyond the scope of this chapter (for a detailed description, see ref. 7), but we will highlight treatment of the YAC DNA leading up to microinjection. The purified DNAs should be used as soon as possible, but can be stored for approx 2–4 wk at 4°C without a detectable increase in degradation. Our YAC DNA preparations generally have concentrations of 5–10 ng/µL. Injections at this concentration reduce embryonic survival and produce a greater frequency of transgenics with deleted YAC copies (64) (K. R. Peterson, unpublished observations). The latter problem may be due to shearing of the YAC DNA as this higher-viscosity DNA solution passes through the needle. We dilute our

Production of Transgenic Mice with YACs

447

DNAs to 1–2 ng/µL in injection buffer and filter the solution through a 0.2 µM Acrodisk. The viscosity of the DNA solution is reduced at this concentration, thus facilitating flow through the injection needle. Filtration removes any undigested agarose, preventing blockage in the needle and does not seem to have a detrimental effect on YAC integrity. Transgenic animals are identified by hybridization of slot blots of tail-derived DNA. Our efficiency of transgenesis is 10–14% for 155- and 248-kb YACs. Up to 25% of these contain only intact YAC copies and up to 60% contain both intact and deleted copies (65) (K. R. Peterson, unpublished data; see Note 9).

5. Notes 1. Batches of ES-qualified serum are generally tested by individual laboratories for the ability to promote high plating efficiency and undifferentiated growth of ES cell lines (27). 2. The optimal concentration of G418 should be determined for each particular ES cell line. 3. Yeast should be grown using a media-to-flask ratio of 1:5 to 1:10 to ensure proper aeration and growth. 4. Inoculating fresh yeast cultures allows for most optimal growth of the yeast and preparation of spheroplasts. Alternatively, yeast cells frozen at 108 cells/mL can be inoculated into selective media and grown for 14 h at 30°C. 5. The exact correlation between cell number and OD600 varies slightly between yeast strains. 6. As yeast spheroplasts lyse in 5% SDS, this method will determine what percentage of cells are spheroplasted. Alternatively, the extent of spheroplasting can be determined by measuring the OD600 of a 1/100 dilution in distilled water. When the OD600 is 1/10 of the starting value, spheroplasting is approx 90% complete. 7. The exact amount of zymolyase and optimal incubation times for the preparation of spheroplasts needs to be determined for each particular yeast strain, cell density, cell volume and enzyme batch. Aliquots of a particular tested batch of zymolyase can be stored at –20°C. 8. The yeast:ES cell ratio can vary from 50:1 to 25:1. 9. A major problem of YAC transgenics is the determination of structural integrity of the YAC transgene copies in selected ES cells and in YAC-containing mice (65). In order to unambiguously demonstrate YAC integrity, a variety of experimental approaches should be undertaken. A combination of PCR, PFGE and conventional Southern-blot hybridization using probes (including species-specific repetitive elements) spanning the YAC is informative, but does not definitively prove the continuity of individual YAC copies, unless restriction sites are chosen that flank the transgene or insert DNA cloned in the pYAC vector. Such sites may be utilized if a restriction map is available or, alternately, unique restriction sites may be introduced at the insert–vector junctions of the YAC prior its use in producing transgenics (66). RecA-assisted restriction endonuclease (RARE) analysis may also be used to visualize entire inserts, but this approach is technically demanding and expensive for routine analysis of large numbers of transgenic lines (67). Finally, modifications of fluorescence in situ hybridization (FISH) may be utilized to examine the orgnization of YAC transgene insertion sites (68), although this approach requires expertise with FISH, as well as multiple FISH probes spanning the YAC. 10. The exact amount of buffer B to be utilized should be empirically determined as the conditions vary considerably for different yeast strains and/or YACs. If the plugs are too concentrated, the YACs will not be separable by PFGE, whereas if the plugs are too dilute, the final purified YAC DNA preparation will not be concentrated enough for efficient lipofection. 11. Alternatively, high-molecular-weight DNA plugs can also be formed in plug molds (see Subheading 4.2., item 2).

448

Jakobovits, Lamb, and Peterson

12. Plugs can be left in solution B O/N, although we have generally obtained the best results when incubated for only 2 h. 13. High-molecular-weight DNA plugs look best by PFGE if allowed to remain in fresh solution E O/N at 4°C. 14. The exact amount of molten plugs to add to the well is determined empirically, but is generally within 3–3.5 mL. It is imperative that the plugs are completely melted before loading the preparative gel, as incompletely melted plugs will result in streaks and unevenness in the gel. The preparative PFGE gel can also be loaded by carefully pouring the molten plugs into the well. The sample will not leak from the large, exapanded well unless the gel is torn. 15. It is convenient to maintain a carboy of distilled water in a 4°C cold room to prepare the 0.5X TBE. 16. Although the exact parameters for YAC DNA isolation are determined empirically, approximate guidelines are as follows: for a 250-kb YAC, a 25-s switch for 29 h; for a 450-kb YAC, a 45-s switch for 37 h; for a 650-kb YAC, a 60 sec switch for 48 h; and for a 1000-kb YAC, a 85-s switch for 49 h. 17. Resolution of PFGE gels will improve with a destaining O/N at 4°C. 18. Alternate methods of YAC DNA concentration include low-speed ultrafiltration (23,35,62) and dialysis with sucrose (36) following agarase digestion of the gel slice. 19. Alternate buffers for maintaining the integrity of high-molecular-weight DNA contain other positively charged molecules, including high salt (62), poly-L-lysine (28–30), and hexamine cobalt chloride (69). 20. It is imperative to completely melt the agarose or the subsequent agarase and lipid-mediated transfection steps will not proceed efficiently. 21. If large amounts of agarose remain undigested at this step, the agarase treatment can be performed a second time. 22. Gibco-BRL recommends performing lipid-mediated transfections in the absence of antibiotics (penicillin–streptomycin). ES media containing penicillin-streptomycin is added the morning after the transfection to protect against possible contamination of the DNA with bacteria. 23. Numerous lipid-based transfection reagents are available on the market. Lipofectin (Gibco-BRL) has provided the most consistent results in delivering YAC DNA prepared as described above into ES cells (K. A. Bardel, L. A. Shaprio, and B. T. Lamb, unpublished observations), although not all of the commercially available lipids have been tested. 24. Kilobase- to megabase-size YAC DNA is susceptible to mechanical shear and enzymatic degradation. To ensure recovery of intact YAC DNA for microinjection, care must be taken to maintain sterility and minimize photodamage and shearing forces. High-quality water (Milli-Q-filtered, Millipore, Bedford, MA) should be used in the preparation of all solutions, buffers, and gels. When possible, solutions should be sterilized by autoclaving or filtration, or made from sterile stock solutions. 25. Bio-Rad plug molds shape 10 plugs of approx 0.25 mL each. Thus, each mold can accommodate 2.5 mL of a yeast cell-agarose suspension. 26. 1900 rpm in a Sorvall GSA rotor. 27. 1770 rpm in a Sorvall HS-4 rotor. 28. Adjust volume of solution, if necessary, so that a ratio of 8 mL solution/mL plug is maintained. 29. The preparative well should be placed as close to the center of the gel as possible to avoid anomalous migration. If DNA migration is uneven, incomplete YAC DNA excision will result, reducing DNA concentration for microinjection. Two YACs of the same or different size may be purified on one gel using separate preparative wells and the PFGE conditions described here.

Production of Transgenic Mice with YACs

449

30. The PFGE conditions given in this protocol are generally applicable for purification of most YACs. The 0.5% agarose MP will fractionate DNAs in the 150-kb to 1.6-Mb size range (63). Alternate agaroses, gel concentrations, and electrophoresis conditions may be utilized to optimize separation and resolution of YACs from yeast chromosomes. These conditions may be determined empirically or Bio-Rad technical services can suggest conditions tailored to maximally fractionate a given size range for isolation of a given size YAC. For example, we routinely use 1% Seakem GTG agarose, 0.5X TBE, 14°C, 200 V, 14-s switch for 24 h to purify 155- and 248-kb YACs. 31. Alternate methods of YAC DNA concentration include low-speed ultrafiltration and dialysis with sucrose following agarase digestion of the PFGE gel slice. 32. If other PFGE conditions are utilized, it may take longer for the YAC DNA to completely migrate from the PFGE gel slice into the Nusieve gel. Use of multiple marker lanes is recommended in this case to monitor the run and determine when it is complete. 33. Do not include any of the PFGE gel slice. Agarase will not digest standard (non-lowmelting point) agarose. 34. It is imperative that high ionic strength be maintained to prevent breakage of YAC DNA at the high temperatures required for agarase treatment and during passage of the DNA through the microinjection needle (35,62). YAC DNA prepared in the presence of 100 mM NaCl is sufficiently resistant to shear (62), although polyamines, such as spermine and spermidine, often are included as well (54,70). Not enough data exist to suggest that the addition of polyamines affords any further degree of resistance to shearing than with high salt alone.

Acknowledgments The authors wish to thank the following for their participation in the development of these protocols: S. Klapholz (A. J.); K. A. Bardel, L. M. Call, S. R. Chideya, and L. S. Kulame (B. T. L.); and A. Gnirke, C. Huxley, C. H. Clegg, and H. S. Haugen (K. R. P.). The development of these protocols was supported in part by the grants from the National Institutes of Health, including AG08012 and AG14451 (B. T. L.), and DK45365 and HL53750 (K. R. P.). References 1. Wagner, T. E., Hoppe, P. C., Jollick, J. D., Scholl, D. R., Hodinka, R. L., and Gault, J. B. (1981) Microinjection of a rabbit β–globin gene into zygotes and its subsequent expression in adult mice and their offspring. Proc. Natl. Acad. Sci. USA 78, 6376–6380. 2. Wagner, E. F., Stewart, T. A., and Mintz, B. (1981) The human β–globin gene and a functional thymidine kinase in developing mice. Proc. Natl. Acad. Sci. USA 78, 5016–5020. 3. Harbers, K., Jahner, D., and Jaenisch, R. (1981) Microinjection of cloned retroviral genomes into mouse zygotes: integration and expression in the animal. Nature 293, 540–542. 4. Gordon, J. W., Scangos, G. A., Plotkin, D. J., Barbosa, J. A., and Ruddle, F. H. (1980) Genetic transformation of mouse embryos by microinjection of purified DNA. Proc. Natl. Acad. Sci. USA 77, 7380–7384. 5. Brinster, R. L., Chen, Y., Trumbauer, E., Senera, A. W., Warren, R., and Palmiter, R. D. (1981) Somatic expression of herpes thymidine kinase in mice following injection of a fusion gene into eggs. Cell 27, 223–231. 6. Jaenisch, R. (1988) Transgenic animals. Science 240, 1468–1474. 7. Gordon, J. W. (1993) Production of transgenic mice, in Guide to Techniques in Mouse Development Vol. 225, (Wassarman, P. M. and DePamphilis, M. L., eds.), Academic Press, San Diego, CA, pp. 747–771.

450

Jakobovits, Lamb, and Peterson

8. Brinster, R. L., Allen, J. M., Behringer, R. R., Gelinas, R. E., and Palmiter, R. D. (1988) Introns increase transcriptional efficiency in transgenic mice. Proc. Natl. Acad. Sci. USA 85, 836–840. 9. Choi, T., Huang, C., Gorman, C., and Jaenisch, R. (1991) A generic intron increases gene expression in transgenic mice. Mol. Cell. Biol. 11, 3070–3074. 10. Burke, D., Carle, G., and Olson, M. (1987) Cloning of large segments of exogenous DNA into yeast artificial chromosomes. Science 236, 806–812. 11. Spencer, F., Ketner, G., Connelley, C., and Hieter, P. (1993) Targeted recombination– based cloning and manipulation of large DNA segments in yeast. Methods 5, 161–175. 12. Hieter, P., Connelly, C., Shero, J., McCormick, M. K., Antonarakis, S., Pavan, W., and Reeves, R. (1990) Yeast artificial chromosomes: promises kept and pending, in Genome Analysis Volume I: Genetic and Physical Mapping (Davies, K. E., and Tilghman, S., eds.), Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 83–120. 13. Mendez, M. J., Green, L. L., Corvalan, J. R. F., Jia, X.–C., Maynard–Currie, C. E., Yang, X.–D., et al. (1997) Functional transplant of megabase human immunoglobulin loci recapitulates human antibody response in mice. Nature Genet. 15, 146–156. 14. Lamb, B. T. and Gearhart, J. D. (1995) YAC transgenics and the study of genetics and human disease. Curr. Opin. Genet. Dev. 5, 342–348. 15. Peterson, K. R., Clegg, C. H., Li, Q., and Stamatoyannopoulos, G. (1997) Production of transgenic mice with yeast artificial chromosomes. Trends Genet. 13, 61–66. 16. Jakobovits, A. (1994) Humanizing the mouse genome. Curr. Biol. 4, 761–763. 17. Jakobovits, A., Moore, A. L., Green, L. L., Vergara, G. J., Maynard–Currie, C. E., Austin, H. A., and Klapholz, S. (1993) Germ–line transmission and expression of a human– derived yeast artificial chromosome. Nature 362, 255–258. 18. Mendez, M. J., Abderrahim, H., Noguchi, M., David, N. E., Hardy, M. C., Green, L. L., et al. (1995) Analysis of the structural integrity of YACs comprimising human immunoglobulin genes in yeast and in embryonic stem cells. Genomics 26, 294–307. 19. Green, L. L., Hardy, M. C., Maynard–Currie, C. E., Tsuda, H., Louie, D. M., Mendez, M. J., et al. (1994) Antigen–specific human monoclonal antibodies from mice engineered with human Ig heavy and light chain YACs. Nature Genet. 7, 13–21. 20. Davies, N. P., Rosewell, I. R., Richardson, J. C., Cook, G. P., Neuberger, M. S., Brownstein, B. H., Norris, M. L., and Brüggemann, M. (1993) Creation of mice expressing human antibody light chains by introduction of a yeast artificial chromosome containing the core region of the human immunoglobulin k locus. Biotechnology 11, 911–914. 21. Zweigerdt, R., Braun, T., and Arnold, H.–H. (1997) Faithful expression of the Myf–5 gene during mouse myogenesis requires distant control regions: A transgene approach using yeast artificial chromosomes. Devlop. Biol. 192, 172–180. 22. Guthrie, C., and Fink, G. R. (1991) Guide to Yeast Genetics and Molecular Biology. Academic, San Diego, CA. 23. Lamb, B. T., Sisodia, S. S., Lawler, A. M., Slunt, H. H., Kitt, C. A., Kearns, W. G., Pearson, P. L., Price, D. L., and Gearhart, J. D. (1993) Introduction and expression of the 400 kilobase amyloid precursor protein gene in transgenic mice. Nature Genet. 5, 22–30. 24. Tsuda, H., Maynard–Currie, C. E., Reid, L. H., Yoshida, T., Edamura, K., Maeda, N., Smithies, O., and Jakobovits, A. (1997) Inactivation of the mouse HPRT locus by a 203–bp retroposon insertion and a 55–kb gene–targeted deletion: Establishment of new HPRT– deficient mouse embryonic stem cell lines. Genomics 42, 413–421. 25. Ramirez–Solis, R., Cavis, A. C., and Bradley, A. (1993) Gene targeting in embryonic stem cells. Methods Enzymol. 225, 855–878. 26. Stewart, C. L. (1993) Production of chimeras between embryonic stem cells and embryos, in Guides to Techniques in Mouse Development Vol. 225, (Wassarman, P. M. and DePamphilis, M. L., eds.), Academic, San Diego, CA, pp. 823–855.

Production of Transgenic Mice with YACs

451

27. Abbondanzo, S. J., Gadi, I., and Stewart, C. L. (1993) Derivation of embryonic stem cell lines, in Guides to Techniques in Mouse Development Vol. 225, (Wassarman, P. M. and DePamphilis, M. L., eds.), Academic, San Diego, CA, pp. 803–823. 28. Lee, J. T., Strauss, W. M., Dausman, J. A., and Jaenisch, R. (1996) A 450 kb transgene displays properties of the mammalian X–inactivation center. Cell 86, 83–94. 29. Lee, J. T., and Jaenisch, R. (1996) A method for high efficiency YAC lipofection into murine embryonic stem cells. Nucleic Acids Res. 24, 5054–5055. 30. Strauss, W. M., Dausman, J., Beard, C., Johnson, C., Lawrence, J. B., and Jaenisch, R. (1993) Germ–line transmission of a yeast artificial chromosome spanning the murine Col1A1 ( a1(I)Collagen) locus. Science 259, 1904–1907. 31. Lamb, B. T., Call, L. M., Slunt, H. H., Bardel, K. A., Lawler, A. M., Eckman, C. B., Younkin, S. G., Holtz, G., Wagner, S. L., Price, D. L., Sisodia, S. S., and Gearhart, J. D. (1997) Altered metabolism of familial Alzheimer’s disease–linked amyloid precursor protein variants in yeast artificial chromosome transgenic mice. Hum. Mol. Genet. 6, 1535–1541. 32. Pearson, B. E., and Choi, T. K. (1993) Expression of the human b–amyloid precursor protein gene from a yeast artificial chromosome in transgenic mice. Proc. Natl. Acad. Sci. USA 90, 10578–10582. 33. Choi, T. K., Hollenbach, P. W., Pearson, B. E., Ueda, R. M., Weddel, G. N., Kurahara, C. G., Woodhouse, C. S., Kay, R. M., and Loring, J. F. (1993) Transgenic mice containing a human heavy chain immunoglobulin gene fragment cloned in a yeast artificial chromosome. Nature Genet. 4, 117–123. 34. Schedl, A., Larin, Z., Montoliu, L., Thies, E., Kelsey, G., Lehrach, H., and Schütz, G. (1993) A method for the generation of YAC transgenic mice by pronuclear microinjection. Nucleic Acids Res. 21, 4783–4787. 35. Peterson, K. R., Clegg, C. H., Huxley, C., Josephson, B. M., Haugen, H. S., Furukawa, T., and Stamatoyannopoulos, G. (1993) Transgenic mice containing a 248–kb yeast artificial chromosome carrying the human β–globin locus display proper developmental control of human globin genes. Proc. Natl. Acad. Sci. USA 90, 7593–7597. 36. Gaensler, K. M. L., Kitamura, M., and Kan, Y. W. (1993) Germ–line transmission and developmental regulation of a 150–kb yeast artificial chromosome containing the human β–globin locus in transgenic mice. Proc. Natl. Acad. Sci. USA 90, 11381–11385. 37. Antoch, M. P., Song, E.–J., Chang, A.–M., Vitaterna, M. H., Zhao, Y., Wilsbacher, L. D., Sangoram, A. M., King, D. P., Pinto, L. H., and Takahashi, J. S. (1997) Functional identification of the mouse circadian Clock gene by transgenic BAC rescue. Cell 89, 655–667. 38. Bungert, J., Dave, U., Lim, K.–C., Lieuw, K. H., Shavit, J. A., Liu, Q., and Engel, J. D. (1995) Synergistic regulation of human β–globin gene switching by locus control region elements HS3 and HS4. Genes Dev. 9, 3083–3096. 39. Frazer, K. A., Narla, G., Zhang, J. L., and Rubin, E. M. (1995) The apolipoprotein(a) gene is regulated by sex hormones and acute–phase inducers in YAC transgenic mice. Nature Genet. 9, 424–431. 40. Frazer, K. A., Ueda, Y., Zhu, Y., Gifford, V. R., Garofalo, M. R., Mohandas, N., et al. (1997) Computational and biological analysis of 680 kb of DNA from the human 5q31 cytokine gene cluster region. Genome Res. 7, 495–512. 41. Hamilton, B. A., Smith, D. J., Mueller, K. L., Kerrebrock, A. W., Bronson, R. T., Berkel, V. V., et al. (1997) The vibrator mutation causes neurodegeneration via reduced expression of PITPa: positional complementation cloning and extragenic suppression. Neuron 18, 711–722. 42. Heard, E., Kress, C., Mongelard, F., Courtier, B., Rougeulle, C., Ashworth, A., et al. (1996) Transgenic mice carrying an Xist–containing YAC. Hum. Mol. Genet. 5, 441–450.

452

Jakobovits, Lamb, and Peterson

43. Hiemisch, H., Schutz, G., and Kaestner, K. H. (1997) Transcriptional regulation in endoderm development: characterization of an enhancer controlling Hnf3g expression by transgenesis and targeted mutagenesis. EMBO J. 16, 3995–4006. 44. Hodgson, J. G., Smith, D. J., McCutcheon, K., Koide, H. B., Nishiyama, K., Dinulos, et al. (1996) Human huntingtin derived from YAC transgenes compensates for loss of murine huntingtin by rescue of the embryonic lethal phenotype. Hum. Mol. Genet. 5, 1875–1885. 45. Huxley, C., Passage, E., Manson, A., Putzu, G., Figarella–Branger, D., Pellissier, J. F., and Fontes, M. (1996) Construction of a mouse model of Charcot–Marie–Tooth disease type 1A by pronuclear injection of human YAC DNA. Hum. Mol. Genet. 5, 563–569. 46. Liu, Q., Bungert, J., and Engel, J. D. (1997) Mutation of gene–proximal regulatory elements disrupts human ε–, γ–, and β–globin expression in yeast artificial chromsome transgenic mice. Proc. Natl. Acad. Sci. USA 94, 169–174. 47. Manson, A. L., Trezise, A. E., MacVinish, L. J., Kasschau, K. D., Birchall, N., Episkopou, V., Vassaux, G., Evans, M. J., Colledge, W. H., Cuthbert, A. W., and Huxley, C. (1997) Complementation of null CF mice with a human CFTR YAC transgene. EMBO J. 16, 4238–4249. 48. Matsuura, S., Episkopou, V., Hamvas, R., and Brown, S. D. M. (1996) Xist expression from an Xist YAC transgene carried on the mouse Y chromosome. Hum. Molec. Genet. 5, 451–459. 49. McCormick, S. P. A., Ng, J. K., Taylor, S., Flynn, L. M., Hammer, R. E., and Young, S. G. (1995) Mutagenesis of the human apolipoprotein B gene in a yeast artificial chromosome reveals the site of attachment for apolipoprotein(a) Proc. Natl. Acad. Sci. USA 92, 10,147–10,151. 50. McCormick, S. P., Ng, J. K., Cham, C. M., Taylor, S., Marcovina, S. M., Segrest, J. P., Hammer, R. E., and Young, S. G. (1997) Transgenic mice expressing human ApoB95 and ApoB97. Evidence that sequences within the carboxyl–terminal portion of human apoB100 are important for the assembly of lipoprotein. J. Biol. Chem. 272, 23616–23622. 51. Montoliu, L., Schedl, A., Kelsey, G., Zentgraf, H., Lichter, P., and Schutz, G. (1994) Germ line transmission of yeast artificial chromosomes in transgenic mice. Reprod. Fertil. Dev. 6, 577–584. 52. Peterson, K. R., Li, Q. L., Clegg, C. H., Furukawa, T., Navas, P. A., Norton, E. J., Kimbrough, T. G., and Stamatoyannopoulos. (1995) Use of yeast artificial chromosomes (YACs) in studies of mammalian development: Production of β–globin locus YAC mice carrying human globin developmental mutants. Proc. Natl. Acad. Sci. USA 92, 5655–5659. 53. Schedl, A., Beermann, F., Thies, E., Montoliu, L., Kelsey, G., and Schütz, G. (1992) Transgenic mice generated by pronuclear injection of a yeast artificial chromosome. Nucleic Acids Res. 20, 3073–3077. 54. Schedl, A., Montoliu, L., Kelsey, G., and Schütz, G. (1993) A yeast artificial chromosome covering the tyrosinase gene confers copy number–dependent expression in transgenic mice. Nature 362, 258–261. 55. Schedl, A., Ross, A., Lee, M., Engelkamp, D., Rashbass, P., Heyningen, V. V., and Hastie, N. D. (1996) Influence of PAX6 gene dosage on development: overexpression causes severe eye abnormalities. Cell 86, 71–82. 56. Smith, D. J., Zhu, Y., Zhang, J., Cheng, J.–F., and Rubin, E. M. (1995) Construction of a panel of transgenic mice containing a contiguous 2–Mb set of YAC/P1 clones from human chromosome 21q22.2. Genomics 27, 425–434. 57. Ainscough, J. F.–X., Koide, T., Tada, M., Barton, S., and Surani, M. A. (1997) Imprinting of Igf2 and H19 from a 130 kb YAC transgene. Development 124, 3621–3632. 58. Yang, W. Y., Model, P., and Heintz, N. (1997) Homologous recombination based modification in Esherichia coli and germline transmission in transgenic mice of a bacterial artificial chromosome. Nature Biotech. 15, 859–865. 59. Mullins, L. J., Kotelevtseva, N., Boyd, A. C., and Mullins, J. J. (1997) Efficient Cre–lox linearisation of BACs: applications to physical mapping and generation of transgenic animals. Nucleic Acids Res. 25, 2539–2540.

Production of Transgenic Mice with YACs

453

60. Nielsen, L. B., McCormick, S. P. A., Pierotti, V., Tam, C., Gunn, M. D., Shizuya, H., and Young, S. G. (1997) Human apolipoprotein B transgenic mice generated with 207– and 145–kilobase pair bacterial artificial chromosomes. J. Biol. Chem. 272, 29752–29758. 61. Wutz, A., Smrzka, O. W., Schweifer, N., Schellander, K., Wagner, E. F., and Barlow, D. P. (1997) Imprinted expression of the Igf2r gene depends on an intronic CpG island. Nature 389, 745–749. 62. Gnirke, A., Huxley, C., Peterson, K., and Olson, M. V. (1993) Microinjection of intact 200– to 500–kb fragments of YAC DNA into mammalian cells. Genomics 15, 659–667. 63. Maule, J. C., Porteous, D. J., and Brookes, A. J. (1994) An improved method for recovering intact pulsed field gel purified DNA of at least 1.6 megabases. Nucleic Acids Res. 22, 3245–3246. 64. Brinster, R. L., Chen, H. Y., Trumbauer, M. E., Yagle, M. K., and Palmiter, R. D. (1985) Factors affecting the efficiency of introducing foreign DNA into mice by microinjecting eggs. Proc. Natl. Acad. Sci. USA 82, 4438–4442. 65. Peterson, K. R. (1997) Production and analysis of transgenic mice containing yeast artificial chromosomes, in Genetic Engineering (Setlow, J. K., ed.), Plenum Press, New York, pp. 235–255. 66. Fairhead, C., Heard, E., Arnaud, D., Avner, P., and Dujon, B. (1995) Insertion of unique sites into YAC arms for rapid physical analysis following YAC transfer into mammalian cells. Nucleic Acids Res. 23, 4011–4012. 67. Ferrin, L. J. and Camerini–Otero, R. D. (1991) Selective cleavage of human DNA: RecA– assisted restriction endonuclease (RARE) cleavage. Science 254, 1494–1497. 68. Rosenberg, C., Voltz, A. K., Lawler, A. L., Lamb, B. T., Stetten, G., and Gearhart, J. D. (1996) Alterations of yeast artificial chromosome transgenic sequences in stretched embyronic stem–cell chromatin visualized by fluorescence in situ hybridization. Cyotogenet. Cell. Genet. 75, 67–70. 69. Kovacic, R. T., Comai, L., and Bendich, A. J. (1995) Protection of megabase DNA from shearing. Nucl. Acids Res. 23, 3999–4000. 70. Couto, L. B., Spangler, E. A., and Rubin, E. M. (1989) A method for the preparative isolation and concentration of intact yeast artificial chromosomes. Nucleic Acids Res. 17, 8010. 71. Lamb, B. T., Bardel, K. A., Kulame, L. S., Anderson, J. J., Holtz, G., Wagner, S. L., Sisodia, S. S., and Hoeger, E. J. (1999) Amyloid production and deposition in mutant amyloid precursor protein and presenilin-1 yeast artificial chromosome transgenic mice. Nature Neurosci. 2, 695–697.

Gene-Targeting Strategies

455

43 Gene-Targeting Strategies Siew-Sim Cheah and Richard R. Behringer 1. Introduction Gene targeting in mouse embryonic stem (ES) cells has become a routine methodology to study gene function in vivo (1). Indeed, gene-targeting core facilities have been established at numerous institutions to facilitate the generation of targeted ES cell lines and the production of mouse chimeras. Although these facilities have centralized many of the unique skills and reagents that are required to generate “knockout” mice, the investigator must still generate a thoughtful strategy to mutate the gene of interest and to subsequently characterize the resulting mutant mice. For the novice, the details that are required to design the optimal strategy to generate a targeted mutation are usually not considered because there are so many targeting strategies that are possible (2). The novice often becomes dazzled by these options and chooses strategies that are highly complex and therefore likely to fail. In addition, beginners often become so excited about generating their first knockout mutation that they rush to generate their gene-targeting vector without carefully considering all of the consequences. Inevitably, these shortcuts come back to haunt them, resulting in unnecessary and costly delays, and many times, starting over to retarget the locus using a different strategy. By investing the time in a thoughtful design of a gene-targeting strategy, one actually saves valuable time and money at subsequent stages of the experiment. In this chapter, we provide a fundamental strategy to generate a null allele for a protein-coding gene by gene targeting in ES cells.

2. General Gene-Targeting Strategy Typically, the first desired mutation to generate in a protein-coding gene is a null allele. There are generally two situations to consider. In the first situation, the proteincoding exons of your gene span a relatively small genomic distance (20 kb or less). In the second situation, the protein-coding exons of your gene span a relatively large distance (greater than 20 kb). For a small gene, the gene-targeting strategy is simple, delete all of the protein coding exons (3–5). Using this strategy, you guarantee the generation of a null allele because it is impossible to generate a protein product from the targeted locus (Fig. 1). In this case, once you have verified that the DNA coding sequences are deleted, it is not necessary to perform mRNA or protein analyses for the From: Methods in Molecular Biology, Vol. 136: Developmental Biology Protocols, Vol. II Edited by: R. S. Tuan and C. W. Lo © Humana Press Inc., Totowa, NJ

455

456

Cheah and Behringer

Fig. 1. Deletion of all protein coding exons for a small gene. The example shown is for the generation of a null allele for Lim1 (5). All of the protein coding exons (open boxes) were replaced by a neomycin (neo) cassette to ensure the generation of a null allele. In this case, the total amount of homology (5.5 kb) was divided as a 1.2-kb 5' arm and a 4.3-kb 3' arm (thick lines). A thymidine kinase (tk) cassette was placed adjacent to the 5' arm of homology. The site of vector linearization is indicated (*). In this situation, 5' and 3' external probes were used to identify targeting events. The sizes of the wild-type (WT) and mutant bands for Southern analysis are indicated. E, EcoRI; RV, EcoRV.

deleted gene. Thus, your time can be allocated for phenotypic characterization studies. If coding sequences remain, mRNA and protein studies of the mutated gene can sometimes be difficult and confusing (6,7). In many situations, biochemical or molecular biological studies have identified domains that are essential for the activity of a specific protein in vitro. This leads some to only delete specific domains because they believe that by doing so, any partial protein that is generated would have no function. This is a fundamental error in logic because partial protein products may have unknown functions. For a larger protein-coding gene, this logic does have some validity because of the constraints imposed by targeting a large gene. However, for small genes, we strongly urge that all protein-coding exons be deleted. Mutating a larger gene requires more thought because simple gene targeting strategies are less efficient for generating very large deletions (>20 kb) (8–10) and may require sophisticated gene-targeting skills that the novice has yet to develop (11,12). In addition, large deletions may remove other genes or regulatory sequences of neighboring genes that reside within large genes. The general strategy we suggest is to generate up to a 20-kb deletion to remove as many of the protein-coding exons as possible, including the exon containing the translation initiation codon. We also suggest that the transcription initiation site, if known, also be included in the targeted deletion (Fig. 2). Although this does not guarantee the generation of a null allele, it increases its likelihood (13,14). The general strategy we describe uses a replacement gene-targeting vector (15). The standard features of a replacement vector are a plasmid backbone containing a positive

Gene-Targeting Strategies

457

Fig. 2. Deletion of the initial protein coding exons for a large gene. The example shown is for the Bmpr locus (13). The coding exons (open boxes) of this gene span 38 kb (top). Therefore, the first two coding exons were deleted. This region (6.3 kb) was replaced by a neo cassette. A total of 9.0 kb of homology was divided as a 3.0-kb 5' arm and a 6.0-kb 3' arm. A tk cassette was placed on the 5' arm of homology. The site of vector linearization is indicated (*). A 3' external probe was used for the initial identification of homologous recombinants. Once these clones were identified they were expanded for subsequent analysis with an internal probe to characterize the structure of the 5' recombination event. The sizes of the wild-type (WT) and mutant bands for Southern analysis are indicated. Bg, BglI; N, NcoI; (RI), EcoRI from phage multicloning site; RV, EcoRV.

selection cassette placed between two regions of chromosomal homology and a negative selection cassette adjacent to one of the homologous arms (Fig. 3). A positive/ negative selection scheme is employed to enrich for homologous recombinants (16). For a replacement strategy, the gene-targeting vector is linearized outside of the regions of homology before introduction into mouse ES cells by electroporation. Drug-resistant ES cell colonies are placed into 96-well tissue-culture plates for expansion and analysis by Southern blot (17). Once the correctly targeted ES cell clones have been identified, they are used for the generation of mouse chimeras (18) or for in vitro differentiation (19) or the generation of teratomas (20). 3. Method 3.1. Isolate Multiple Clones for the Gene from a Mouse Genomic Library In constructing your targeting vector, it is important to use genomic DNA that is isogenic to the mouse ES cell line that will be used for the gene-targeting experiments to facilitate the maximum frequency of homologous recombination events (21). For

458

Cheah and Behringer

Fig. 3. General design of a replacement gene-targeting vector. The floxed neo and tk cassettes can be placed in either forward or reverse orientation relative to the orientation of transcription for the gene to be targeting. Mutant allele 1 represents the targeted locus with neo and Mutant allele 2 represents the targeted locus that has had the neo, cassette removed by cre recombinase. The sizes of the wild-type (WT) and mutant bands for Southern analysis are indicated.

historical reasons, the majority of mouse ES cell lines are derived from 129 inbred mouse strains (22). It is essential to note that there are many substrains of 129 that can be quite genetically diverse, especially the 129/SvJ substrain (23). Therefore, it is important to determine the precise strain and substrain of mouse used to generate the ES cell line that you will use in your gene-targeting experiments and screen a genomic library from that same strain/substrain. Currently, one very popular mouse ES cell line called R1 is derived from a (129/Sv × 129/SvJ)F1 embryo (24). The 129 strain mouse genomic libraries are commercially available from numerous sources (Stratagene, La Jolla, CA; Genome Systems, St. Louis, MO; Research Genetics, Huntsville, AL). It is important to isolate multiple genomic clones for your gene, especially if it is a large gene. This will provide more genomic sequences and, thus, more options in designing a targeting vector. In addition, be aware that genomic libraries may contain single clones that contain DNA fragments from different regions of the genome. The coligated genomic inserts in these clones can eliminate all chances of generating or recognizing a targeted mutation. Generate a detailed restriction map for each genomic clone and verify that this map truly matches the chromosomal locus by Southern-blot

Gene-Targeting Strategies

459

hybridization of isogenic genomic DNA derived from cells or tissues. There have been many instances when correct gene-targeting had been obtained, but results of the Southern analysis were not as predicted because of an erroneous restriction map.

3.2. Replacement Vector Design Once you have obtained and characterized genomic clones for the gene of interest, you must design your gene-targeting strategy “on paper.” We urge beginners to develop a gene-targeting strategy that employs a Southern-blot analysis to identify targeting events. The 96-well tissue-culture plate method for the rapid screening of ES cell clones by Southern analysis is exceedingly easy for beginners to master very quickly (17). Do not use a polymerase chain reaction (PCR) genotyping strategy because it imposes constraints on your gene-targeting vector design. In addition, we have found that PCR genotyping is more problematic for beginners than Southern-blot analysis. Furthermore, PCR results must always be confirmed by Southern analysis. In your design, identify genomic regions that can serve as Southern blot probes to recognize targeting events. Do not generate your targeting vector until your potential probes have been verified on Southern blots of isogenic genomic DNA. Once the probes have been verified, then construct your gene-targeting vector as shown in Fig. 3. The details of each aspect of the replacement vector are discussed in the following.

3.2.1. Choice of Homologous Regions for the Targeting Vector The total amount of chromosomal homology should be between 5 and 8 kb (15). Smaller amounts of homology may reduce the targeting frequency and significantly larger amounts of homology become unwieldy for targeting vector construction and homologous recombinant identification. The amount of homologous sequences for each arm of homology in the vector should be about half of the total amount of homology. For example, a vector with a total of 6 kb of homology may have two arms of homologous sequence that are approx 3 kb each. Each arm of homology in the targeting vector should flank the protein-coding exons and other sequences that you wish to delete. Remember that neighboring genes may lie very close to your gene of interest (25). Therefore, try not to delete beyond the known sequences of your gene.

3.2.2. Positive Selection Cassette We suggest that you use a neomycin phosphotransferase (neo) gene expression cassette for selection with G418 that has the mouse phosphoglycerate kinase (PGK) promoter and the polyadenylation signal from the bovine growth hormone (bpA) gene (PGKneobpA) because it is more efficient than other neo expression cassettes that are typically used in gene-targeting experiments (26). Although there are other positive selectable markers that can be used, (e.g., hygromycin, puromycin, etc.), we suggest that the beginner use neomycin because it is routinely used by most laboratories and most readily available feeder cell lines, on which ES cells grow, are G418 resistant. The neo cassette can be placed in either forward or reverse orientation relative to the direction of transcription of your gene. The presence of a selectable marker cassette with its exogenous promoter has been documented to alter gene expression at targeted loci (27). Therefore, the neo cassette in the strategy we outline should be flanked by loxP sequences (floxed) to provide the

460

Cheah and Behringer

option of subsequently removing the neo cassette with cre recombinase (28). You lose nothing by cloning a floxed–neo expression cassette into your gene-targeting vector and gain many potential benefits. The option to remove the neo cassette using simple methods can potentially save effort and the costs of regenerating another mutant should any question arise about the altered transcription of your gene or neighboring genes (29). The floxed–neo cassette strategy also provides the opportunity to remove the selectable marker cassette from the targeted locus in vitro for subsequent retargeting of the remaining wild-type allele with the original targeting vector and drug selection to generate homozygous mutant ES cell lines, a process called “marker recycling” (30). The floxed–neo cassette strategy also allows one to easily generate two different alleles for your gene of interest, one with neo and one without, that can be distinguished by Southern blot or PCR. The availability of two different alleles can be exploited for potential chimera experiments (31,32). The floxed–neo cassette can be removed in ES cells by transient cre expression in vitro (30) by the injection of a cre expression vector into the pronuclei of eggs containing the floxed gene (33) or by simply crossing mice carrying the floxed gene with cre expressing transgenic mice (34).

3.2.3. Negative Selection Cassette We suggest that you use a herpes simplex virus thymidine kinase gene expression cassette for negative selection with gancyclovir or FIAU (1-2-deoxy-2-fluroro1-β-D-arabinofuranosyl-5-iodouracil). The MC1TKpA vector works very well with FIAU, providing a five- to 10-fold enrichment for targeting events (35). An alternate and convenient negative selection cassette is one that expresses the diphtheria toxin A-chain (DT-A) gene (36). The advantage of the DT-A strategy is that it is not necessary to add any drug selection to the culture medium because expression of DT-A in cells is toxic. The beginner should note that even after positive/negative drug selection, many of the resulting ES cell colonies are nontargeted (random integration) events. This occurs because the negative selection cassette used in the gene-targeting vector is not expressed, possibly because of damage or integration into chromosomal regions that inhibit its expression. The negative selection cassette can be placed in either forward or reverse orientation relative to the direction of transcription of your gene.

3.2.4. Site of Targeting Vector Linerization Gene-targeting vectors are routinely introduced into ES cells by electroporation in a linearized form. Thus, in the design of your targeting vector, a unique restriction enzyme site must be present as a site for linearization. For a replacement gene-targeting vector, the linearization site must be outside of the regions of homology. Typically, the linearization site is located at the junction of one of the homologous arms and the plasmid backbone. We suggest that the linearization site should not be at the junction between the negative selection cassette and the plasmid backbone (Fig. 3) to reduce the chances of the negative selection cassette from being degraded by nucleases after introduction into cells, causing an increase in the background of nontargeted clones.

3.2.5. Restriction Enzyme Sites for Southern Analysis To identify homologous recombinants, Southern analysis using a diagnostic restriction enzyme digest with a probe that is external to the regions of homology

Gene-Targeting Strategies

461

(external probe) included in the targeting vector must be performed. Therefore, it is important to consider restriction enzyme sites that can be used to differentiate between targeted and nontargeted events during the construction of the targeting vector. A unique restriction enzyme site should be introduced by the positive selection cassette. Remember that this restriction enzyme site should be located outside of the floxed–neo expression cassette so that after cre recombinase expression, the diagnostic restriction enzyme site will still remain. In this way, a diagnostic digest will yield a smaller DNA fragment for the mutant allele in comparison to the wild-type allele that can be recognized by the external probe. This is essential because partial restriction enzyme digestions can occur in the cruder DNA samples that are prepared using the 96-well tissue-culture plate method for the rapid screening of recombinant ES cells. Thus, if the mutant allele yields a DNA fragment that is larger than the wild-type allele, a partial restriction enzyme digest can cause difficulties during the Southern analysis. We suggest choosing restriction enzymes that work efficiently on these cruder DNA preparations for the most consistent results for genomic Southerns. The restriction enzymes that have worked well for us include BamHI, BglI, BglII, EcoRI, EcoRV, PstI, and SstI, whereas KpnI, XbaI, XhoI, and XmnI have proven to be problematic.

3.2.6. 5' and 3' External Probes for the Identification of Targeted ES Cell Clones It is important to use both 5' and 3' external probes that can recognize the structure of the gene targeting events on the 5' and 3' arms of homology. It is not uncommon for correct targeting to be obtained on one arm of homology but not the other. If both 5' and 3' external probes cannot be found, the initial targeting event can be identified using one external probe that confirms the structure of the targeting event on one side of homology. Once this subset of targeted ES cell clones is identified, a probe within the region of targeting vector homology (internal probe, for example neo) can be used to verify the structure of the targeting event on the other side of homology. Do not use the PGK promoter sequences of the neo expression cassette as an internal probe because it is derived from mouse and will recognize the endogenous mouse PGK gene. 4. Summary Gene targeting in mouse ES cells is a powerful method for studying gene function in vivo. For the novice, this combination of molecular biology, specialized tissue-culture cell lines, and mouse reproductive biology can be daunting. We present a straightforward, one might say constrained, guide for novices of gene targeting to generate a null allele in large or small protein-coding genes. The method we outline has evolved from years of experience of training and advising beginners on this powerful technology. We believe that a good design for a gene-targeting strategy ultimately saves time, money, and research animal lives. Once you feel comfortable with a fundamental knockout, we suggest you consider the new and exciting gene-targeting variations that can be used to address important biological questions (2). Acknowledgments The authors thank Yuji Mishina and Maki Wakamiya for helpful comments on the manuscript.

462

Cheah and Behringer

References 1. Torres, M. (1998) The use of embryonic stem cells for the genetic manipulation of the mouse. Curr. Top. Dev. Biol. 36, 99–114. 2. Bradley, A. and Liu, P. (1996) Target practice in transgenics. Nature Gen. 14, 121–123. 3. Chen, Z.-F. and Behringer, R. R. (1995) twist is required in head mesenchyme for cranial neural tube morphogenesis. Genes Dev 9, 686–699. 4. Rivera-Pérez, J. A., Mallo, M., Gendron-Maguire, M., Gridley, T., and Behringer, R. R. (1995) Goosecoid is not an essential component of the mouse gastrula organizer but is required for craniofacial and rib development. Development 121, 3005–3012. 5. Shawlot, W. and Behringer, R. R. (1995) Requirement for Lim1 in head-organizer function. Nature 374, 425–430. 6. Moens, C. B., Auerbach, A. B., Conlon, R. A., Joyner, A. L., and Rossant, J. A. (1992) Targeted mutation reveals a role for N-myc in branching morphogenesis in the embryonic mouse lung. Genes Dev. 6, 691–704. 7. Hasty, P., Bradley, A., Morris, J. H., Edmondson, D. G., Venuti, J. M., Olson, E. N., and Klein, W. H. (1993) Muscle deficiency and neonatal death in mice with a targeted mutation in the myogenin gene. Nature 364, 501–506. 8. Mombaerts, P., Clarke, A. R., Hooper, M. L., annd Tonegawa, S. (1991) Creation of a large genomic deletion at the T-cell antigen receptor beta-subunit locus in mouse embryonic stem cells by gene targeting. Proc. Natl. Acad. Sci. USA 88, 3084–3087. 9. Zhang, H., Hasty, P., and Bradley, A. (1994) Targeting frequency for deletion vectors in embryonic stem cells. Mol. Cell. Biol. 14, 2404–2410. 10. Tsuda, H., Maynard-Currie, C. E., Reid, L. H., Yoshida, T., Edamura, K., Maeda, N., Smithies, O., and Jakobovits, A. (1997) Inactivation of the mouse HPRT locus by a 203-bp retroposon insertion and a 55-kb gene-targeted deletion: establishment of new HPRTdeficient mouse embryonic stem cell lines. Genomics 15, 413–421. 11. Ramírez-Solis, R., Liu, P., and Bradley, A. (1995) Chromosome engineering in mice. Nature 378, 720–724. 12. Li, Z. W., Stark, G., Gotz, J., Rulicke, T., Gschwind, M., Huber, G., Muller, U., and Weissmann, C. (1996) Generation of mice with a 200-kb amyloid precursor protein gene deletion by Cre recombinase-mediated site-specific recombination in embryonic stem cells. Proc. Natl. Acad. Sci. USA 93, 6158–6162. 13. Mishina, Y., Suzuki, A., Ueno, N., and Behringer, R. R. (1995) Bmpr encodes a type I bone morphogenetic protein receptor that is essential for gastrulation during mouse embryogenesis. Genes Dev. 9, 3027–3037. 14. Mishina, Y., Rey, R., Finegold, M. J., Matzuk, M. M., Josso, N., Cate, R. L., and Behringer, R. R. (1996) Genetic analysis of the Mullerian-inhibiting substance signal transduction pathway in mammalian sexual differentiation. Genes Dev. 10, 2577–2587. 15. Hasty, P. and Bradley, A. (1993) Gene targeting vectors for mammalian cells, in Gene Targeting: A Practical Approach (Joyner, A. L., ed.), IRL, Oxford, pp. 1–31. 16. Mansour, S. L., Thomas, K. R., and Capecchi, M. R. (1988) Disruption of the protooncogene int-2 in mouse embryo-derived stem cells: a general strategy for targeting mutations to non-selectable genes. Nature 336, 348–352. 17. Ramírez-Solis, R., Rivera-Pérez, J., Wallace, J. D., Wims, M., Zheng, H., and Bradley, A. (1992) Genomic DNA microextraction: a method to screen numerous samples. Analytical Biochem 201, 331–335. 18. Wood, S. A., Allen, N. D., Rossant, J., Auerbach, A., and Nagy, A. (1993) Non-injection methods for the production of embryonic stem cell-embryo chimaeras. Nature 365, 87–89. 19. Robertson, E. J. (ed.) (1987) Embryo-derived stem cell lines, in Teratocarcinomas and Embryonic Stem Cells: A Practical Approach, IRL, Oxford, pp. 71–112.

Gene-Targeting Strategies

463

20. Shaw-White, J. R., Denko, N., Albers, L., Doetschman, T. C., and Stringer, J. R. (1993) Expression of the lacZ gene targeted to the HPRT locus in embryonic stem cells and their derivatives. Transgenic Res. 2, 1–13. 21. te Riele, H., Maandag, E. R., and Berns, A. (1992) Highly efficient gene targeting in embryonic stem cells through homologous recombination with isogenic DNA constructs. Proc. Natl. Acad. Sci. USA 89, 5128–5132. 22. Papaioannou, V. and Johnson, R. (1993) Production of chimeras and genetically defined offspring from targeted ES cells, in Gene Targeting: A Practical Approach (Joyner, A. L., ed.), IRL, Oxford, pp. 1–31. 23. Threadgill, D. W., Yee, D., Matin, A., Nadeau, J. H., and Magnuson, T. (1997) Genealogy of the 129 inbred strains: 129/SvJ is a contaminated inbred strain. Mamm. Genome 8, 390–393. 24. Nagy, A., Rossant, J., Nagy, R., Abramow-Newerly, W., and Roder, J. C. (1993) Derivation of completely cell culture-derived mice from early-passage embryonic stem cells. Proc. Natl. Acad. Sci. USA 90, 8424–8428. 25. Dresser, D. W., Hacker, A., Lovell-Badge, R., Guerrier, D. (1995) The genes for a spliceosome protein (SAP62) and the anti-Mullerian hormone (AMH) are contiguous. Hum. Mol. Genet. 4, 1613–1618. 26. Soriano, P., Montgomery, C., Geske, R., and Bradley, A. (1991) Targeted disruption of the c-src proto-oncogene leads to osteopetrosis in mice. Cell 64, 693–702. 27. Fiering, S., Epner, E., Robinson, K., Zhuang, Y., Teiling, A., Hu, M., Martin, D. I., Enver, T., Ley, T. J., and Groudine, M. (1995) Targeted deletion of 5’HS2 of the murine betaglobin LCR reveals that it is not essential for proper regulation of the beta-globin locus Genes Dev. 9, 2203–2213. 28. Sauer, B. and Henderson, N. (1988) Site-specific DNA recombination in mammalian cells by the CRE recombinase in bacteriophage P1. Proc. Natl. Acad. Sci. USA 85, 5166–5170. 29. Olson, E. N., Arnold, H. H., Rigby, P. W., and Wold, B. J. (1996) Know your neighbors: three phenotypes in null mutants of the myogenic bHLH gene MRF4. Cell 85, 1–4. 30. Abuin, A. and Bradley, A. (1996) Recycling selectable markers in mouse embryonic stem cells. Mol. Cell. Biol. 16, 1851–1856. 31. Quinn, J. C., West, J. D., and Hill, R. E. (1996) Multiple functions for Pax6 in mouse eye and nasal development. Genes Dev. 10, 435–446. 32. Rivera-Pérez, J. A., Wakamiya, M., and Behringer, R. R. (1999) Goosecoid acts cell autonomously for the maintenance of mesenchymal tissues during craniofacial development. Development 126, 3811–3821. 33. Sunaga, S., Maki, K., Komagata, Y., Ikuta, K., and Miyazaki, J. I. (1997) Efficient removal of loxP-flanked DNA sequences in a gene-targeted locus by transient expression of Cre recombinase in fertilized eggs. Mol. Reprod. Dev. 46, 109–113. 34. Lasko, M., Sauer, B., Mosinger, B. Jr, Lee, E. J., Manning, R. W., Yu, S. H., Mulder, K. L., and Westphal, H. (1992) Targeted oncogene activation by site-specific recombination in transgenic mice. Proc. Natl. Acad. Sci. USA 89, 6232–6236. 35. McMahon, A. P. and Bradley, A. (1990) The Wnt-1 (int-1) proto-oncogene is required for development of a large region of the mouse brain. Cell 62, 1073–1085. 36. McCarrick, J. W., Parnes, J. R., Seong, R. H., Solter, D., and Knowles, BB. (1993) Positive-negative selection gene targeting with the diphtheria toxin A-chain gene in mouse embryonic stem cells. Transgenic Res. 2, 183–190.

Chimeric Animals and Germline Transmission

465

44 Chimeric Animals and Germline Transmission Jaspal S. Khillan 1. Introduction Gene targeting by homologous recombination in pleuripotent embryonic stem (ES) cells provides a powerful tool to introduce specific mutations in the genome of intact animal (1–5). It, therefore, allows to unravel the function of genes that control development and differentiation in the organism. Generally, a gene targeting cassette that contains positive–negative selection markers is prepared in which an exon of the target gene is interrupted by the gene for neomycin resistance, which serves as a positive selection marker (6–8). A herpes simplex virus thymidine kinase (HSV TK) gene is fused at either one or both ends of the genomic sequences for negative selection. The targeting cassette is introduced into ES cells, and the cells in which the endogenous gene is disrupted are selected with G418 and gancyclovir or FIAU (6,8). The cells in which the DNA is inserted randomly will die as a result of the incorporation of gancyclovir or FIAU, the thymidine analogs, by the thymidine kinase gene, which blocks the DNA synthesis. The selected ES cell clones are then used to prepare chimeric animals for germline transmission. The chimeric animals are prepared either by direct microinjection of ES cells into the cavity of a blastocyst (1) or by aggregation of ES cells with morula followed by transfer into pseudopregnant females. Various aggregation methods to prepare chimeric animals are described in literature (9–11) (see Note 1). Here I will describe a one-step coculture method recently established in my own laboratory (12). 2. Materials 2.1. Animals 1. Four to five week old C57BL6 or FVB/N female mice as embryo donors and 8- to 10-wkold stud males of the same strains. 2. Five- to six-wk-old female mice of CD1 or NIH-GP strains as pseudopregnant recipient mothers.

2.2. Hormones 1. Pregnant mare serum gonadotropin (PMSG). 2. Human chorionic gonadotropin (HCG). From: Methods in Molecular Biology, Vol. 136: Developmental Biology Protocols, Vol. II Edited by: R. S. Tuan and C. W. Lo © Humana Press Inc., Totowa, NJ

465

466

Khillan

A solution of 50 IU/mL for each hormone is prepared in 0.9% NaCl. One-milliliter aliquots are stored frozen, protected from light up to 4–6 wk (see Note 2).

2.3. Anesthesia Avertin is the most commonly used anesthesia in mice. A stock solution of 100% avertin is prepared by mixing 10 g of tribromo ethanol with 25 mL tertiary amyl alcohol. The stock solution is diluted to 2.5% in water before use and stored in the dark at room temperature for prolonged periods. About 0.015–0.017 mL/g body weight is administered to anesthetize adult mice of 20–25 g weight. 2.4. Embryo Culture Medium A modified Whitten’s medium, M16 medium (13), is used to culture preimplantation embryos. The composition of the medium is shown in Table 1. The medium can be stored up to 2 wk at 4°C. The medium, can also be purchased from Gibco-BRL (Grand Island, NY) (Brinster’s BMOC-3 medium).

2.5. Tissue Culture Medium 1. Blastocyst microinjection medium: Dulbecco’s modified Eagle medium (DMEM) w/o sodium pyruvate, w/o sodium bicarbonate, 10% fetal bovine serum (FBS) and 25 mM HEPES pH 7.2. The osmolarity of the medium should be around 290–300 mosmol. 2. Primary fibroblast (PF) culture medium: This medium contains 10% FBS and 1% penicillin/streptomycin (pen/strep) in DMEM. 3. Embryonic stem cell (ES) medium: The medium is prepared by mixing 90 mL of FBS, 6 mL of 10 mM nonessential amino acids, 3 mL of 100X pen/strep, 1 mL of β-mercaptoethanol, and 60 µL lypmhocyte inhibitory factor (LIF; available from Gibco-BRL as ES-GRO cat. no. 3275SB), which provides 1000 IU/mL, with 500 mL of DMEM. The medium should be used within 2–3 wk (see Note 3). 4. ES cell freezing medium: The 2X freezing medium contains 20% dimethyl sulfoxide (DMSO), 20% FBS, and 60% ES medium. 5. For morula aggregation, the embryo culture medium is supplemented with bovine serum albumin (BSA) to a final concentration of 10 mg/mL.

2.6. Microtools (See Note 4) 1. Isolation and transfer pipet: Glass capillaries (World Precision Instruments, Inc., Sarasota, FL, 1 mm internal diameter [id]) are prepulled on a gas burner. The prepulled capillary is cut with a diamond pencil so that the opening of the needle is about 120–140 µm id. The tip of the capillary is polished on a microforge with a 100- to 120-µm opening. The same pipet is used for transfer of embryo into pseudopregnant females. 2. Preparation of coculture dish: A 35-mm petridish with 8–10 microwells, 0.3–0.5 mm outer diameter (od) at the top and 0.1–0.2 mm at the bottom, are constructed in a circle of 0.5 cm (Fig. 1). The microwells are constructed by pressing the blunt end of a fire polished glass Pasteur pipet against the bottom of a 35-mm Petri dish. The Petri dishes can also be purchased from Genome System Inc. (St. Louis, MO) (see Note 5).

2.7. Equipment 1. 2. 3. 4.

Microforge. CO2 incubator. Stereomicroscope. Surgical instruments.

Chimeric Animals and Germline Transmission

467

Table 1 Embryo Culture Medium Compound NaCl KCl CaCl2·2H2O KH2PO4 MgSO4·7H2O NaHCO3 Sodium lactate Sodium pyruvate Glucose Bovine serum albumin Penicillin G (100 U/mL) streptomycin sulfate (50 µg/mL) Phenol red

mM

g/L

94.66 4.78 1.71 1.19 1.19 25.0 23.28 0.33 5.56

5.533 0.356 0.252 0.162 0.293 2.101 2.610 0.036 1.000 4.000 0.060 0.050 0.010

Note: The aggregation medium is supplemented with BSA to the final concentration of 10 mg/mL.

Fig. 1. A coculture Petri dish. Microwells are constructed on the surface of a 35-mm petridish in a circle of about 0.5 cm diameter. Each Petri dish contains 10 wells with a diameter of about 0.2 mm at the bottom and about 0.5 mm at the top.

2.8. Radiolabeled Compounds 1. A 32 P-labeled deoxyribonucleotide triphosphate to prepare gene-specific probes and testing the recombination event in ES cells and in embryos by Southern-blot analysis (see Note 6).

468

Khillan

3. Methods Chimeric animals can be prepared either by microinjection of ES cells into blastocyst (1) or by aggregation of ES cells with morula (9–12). Following are the steps involved in generating chimeric animals by the one-step aggregation method (12).

3.1. Preparation of Primary Fibroblasts 1. To establish primary fibroblast cells (see Note 3), normal or transgenic females with a neomycin-resistance gene are crossed with respective males. Sacrifice mothers at 14–16 d of pregnancy and remove embryos from the uterine horns. 2. Wash embryos once in HEPES saline and dissect individual embryo to remove head and soft tissues such as liver heart and other viscera. 3. Wash carcass twice in HEPES saline buffer followed by mincing into small pieces and transfer to a clean 150-mL flask. 4. Trypsinize tissue with 30 mL trypsin/EDTA for 30 min with constant stirring. Add additional 30 mL of trypsin/EDTA and stir again for 30 min. 5. Add 60 mL of medium with 10% FBS and dissociate tissue by vigorous pipeting and transfer into 50-mL falcon tubes to allow larger tissue pieces to settle. 6. Transfer supernatant to 100-mm Petri dishes (about one dish for each embryo). After 3–4 d, the confluent cells are harvested. Most of the cells are frozen for future use. Propagate one Petri dish to prepare feeder layer cells. The cells are generally used up to fifth passage because their efficiency to support ES cells declines after the fifth passage.

3.2. Preparation of Feeder Plates 1. To use as feeder layers, PF cells are inactivated either by treating with mitomycin C or by irradiation (5) (see Note 7). The PF cells are irradiated at 4000–6000 rads. The cells are stored in individual vials, 1 × 106 cells/vial, in liquid nitrogen. 2. Treat 60-mm or 100-mm plates with 0.2% gelatin for 2 h. 3. Transfer one vial of frozen PF cells into a 60-mm plate and two vials of cells into a 100-mm plate. 4. Allow cells to settle overnight and change the medium the next day. The cells form a monolayer on the bottom of the Petri dish. 5. Switch over to ES cell medium about 2 h before the transfer of ES cells. The plates can be used up to 10 d after preparation.

3.3. Preparation of Gene Targeting Vector 1. A gene-targeting cassette is constructed in which an exon of the gene is interrupted by neomycin-resistance gene that also serves as a positive selection marker (6–8) (Fig. 2). 2. Typically, about 5–10 kb of the genomic DNA is cloned into a targeting vector in such a way that the neomycin gene is flanked by the genomic sequences. 3. On one or both ends of the genomic sequences, a gene for herpes simplex virus thymidine kinase (HSV TK) is ligated as a negative selection marker. Both of the selector genes are directed by a constitutive promoter that express in ES cells (e.g., phosphoglycero kinase [PGK]). 4. The targeting cassette is then cut out from the vector sequences and purified on agarose gel (Fig. 2).

3.4. Electroporation of DNA into ES Cells 1. Thaw one vial of ES cells on a 60-mm plate with a feeder layer and culture till confluency (see Note 8) (Fig. 3). Trypsinize cells and wash one time with fresh ES cell medium. Resuspend 1–2 × 107 cells in 700 µL of phosphate buffer saline (PBS).

Chimeric Animals and Germline Transmission

469

Fig. 2. The gene targeting cassette (top panel) with gene for neomycin resistance (neo) and herpes simplex thymidine kinase (HSV-TK). The neomycin gene is introduced into one of the exons (black box), whereas the HSV TK gene is ligated outside the region of the homology. The middle panel represents the chromosomal DNA. The bottom panel shows that after homologous recombination, the HSV-TK gene is excluded.

Fig. 3. The culture of ES cells after 48 h of culture. The cells were cultured over the feeder layer. The round morphology of the colonies indicates undifferentiated cells. 2. Transfer cells to a 800 µL electroporation vial and mix 25 µg linearized DNA dissolved in 100 µL of PBS. Electroporate DNA at 280 V and 500 µF capacitance (Bio-Rad Pulsar Electroporator, Bio-Rad, Hercules, CA).

470

Khillan

3. Allow cells to stand at room temperature for 10 min followed by dilution in 5 mL of ES medium. Transfer cells to six 100-mm Petri dishes with feeder layer. 4. After 48 h, the cells are selected by treating with ES cell medium containing 350 µg/mL of G418 and 2 µm of gancylovir or 0.2 µM FIAU. 5. Change medium every day. After about 8–10 d the individual colonies can be seen growing in the petridish (5). 6. The individual colonies are harvested with a yellow tip and trypsinized in 50 µL of trypsin. Trypsinized cells are transferred to a well of 24-well plate and the cells are allowed to grow for 2–3 d. 7. Confluent cells from each well are harvested by trypsinization. One half of the cells are transferred to a new 24-well plate with 2X ES cell freezing medium and the other half is transferred back to the same petridish for further culture. The cells are cultured for 3–5 d to isolate DNA. 8. The plate with freezing medium are stored at –70°C till the screening is complete.

3.5. Screening of ES Clones 1. After the cells in the 24-well plate are confluent, add 100 µL of lysis buffer (10 mM Tris pH 7.5, 10 mM EDTA, 10 mM NaCl, 0.5% sarcosyl, and 1 mg/mL proteinase K) and transfer lysate into a 1.5-mL eppendorf tube and incubate overnight at 55°C. 2. Add equal volume of 1:1 phenol chloroform mix and shake well to mix the contents. Centrifuge and transfer supernatant to another tube and precipitate DNA with an equal volume of isopropanol. 3. Pellet DNA at 13,000g in a microfuge centrifuge. Wash pellet with 70% ethanol and dissolve in 10 mM Tris and 1 mM EDTA buffer. 4. The DNA is analyzed either by PCR or by Southern-blot analysis.

3.6. Preparation of ES Cells for Aggregation with Morula 1. Thaw ES cells in which the gene is knocked out from the frozen 24-well plates and transfer onto 60-mm plates with feeder layer. 2. Change medium daily and harvest cells after 48 h or as soon as the ES cells are confluent by adding 1–2 mL of trypsin/EDTA followed by incubation for 5–8 min at 37°C. Add 10 mL of ES medium . 3. Pipet cells up and down several times to break the lumps. The cells should be dispersed as individual cells. 4. Transfer cells to a 50-mL tube followed by centrifugation at 1000g for 5 min. 5. Resuspend in 10 mL of ES medium and allow to stand for 10 min on ice for feeder cells to settle down. 6. Collect top 5 mL of the cells for the preparation of chimeras. 7. The remaining cells are cultured again on new PF plates for use on following day and a part of the cells are frozen in liquid nitrogen for future use.

3.7. Preparation of Vasectomized Males 1. Anaesthetize adult 6- to 8-wk-old male mice with avertin. Make a horizontal incision in the abdominal region at the level of hind legs. Pull out testis on each side along with the vas deference. 2. Cut about half cm of the vas deference and push organs back into the peritoneal cavity, followed by closing the skin with the wound clips. The testis descend into the scrotal sac automatically. 3. Allow 5–7 d for mice to recover and then mix with the females.

Chimeric Animals and Germline Transmission

471

4. Check females daily for the vaginal plug which is formed from the coagulation of proteins secreted in the semen. The plugs generally fall off after 12–14 h of mating. About 30 vasectomized males are sufficient to provide three to four pseudopregnant females every day.

3.8. Preparation of Pseudopregnant Females 1. Mix two to three CD1 female mice with vasectomized males. 2. Check plugs the next day. 3. Use 2.5-d pseudopregnant females for blastocyst transfer (see Note 9).

3.9. Collection of Morulas 1. Superovulate 8–10 FVB/N or C57BL6 female mice by injecting 5 IU PMSG intraperitoneally. After 48 h, inject 5 IU HCG (see Note 10). 2. Mix female mice with the stud males overnight and check females for vaginal plugs the next morning. Use only the plugged females for embryo isolation. Unplugged females can be recycled after 3–4 wk. 3. The next morning separate plugged females from the vasectomized males as foster mothers (see Note 11). 4. Prepare the following 60-mm Petri dishes with embryo culture medium and incubate at 37°C for about 1 h. First Petri dish—approx 5 mL medium. Second Petri dish—approx 8–10 mL of medium. 5. Sacrifice pregnant females by cervical dislocation and flush abdominal area with 70% ethanol. The females are sacrificed after 2.5 d of mating for morula isolation and after 3.5 d for blastocyst isolation. 6. Open peritoneal cavity, cut out each uterine horn separately, and transfer into the first Petri dish. 7. Using a 3-mL syringe filled with medium, flush uterine horns individually in a fresh 60-mm Petri dish with about 1 mL medium per uterine horn. 8. Collect morulas or blastocysts with the help of an isolation pipet and wash in fresh medium. 9. Transfer embryos to a 35-mm Petri dish with fresh medium and culture in a CO2 incubator at 37°C until further use.

3.10. Preparation of Plates for Aggregation of ES Cells and Morula 1. Transfer ES cells (Subheading 3.4.) to a 1.5-mL microcentrifuge tube. 2. Centrifuge at 1000 rpm in a microcentrifuge for 2 min at room temperature followed by washing two times in 1 mL of morula aggregation medium (see Subheading 2.5., item 5). 3. Resuspend cells in the same medium at a concentration of 1–2 × 104 cells/mL. 4. Transfer 100 µL of ES cell suspension over the microwells and cover with paraffin oil. 5. Allow cells to settle for 10 min at room temperature. About 15–20 cells settle down in each microwell.

3.11. Preparation of Zona-Free Morula and Aggregation with ES Cells: 1. While the cells are settling down in microwell plates, collect 30–50 healthy morulas in a minimum volume of medium. 2. Transfer embryos into a drop of 100 mL acid Tyrode solution (137 mM NaCl; 2.7 mM KCl; 0.5 mM MgCl2·6H2O; 5.6 mM glucose; 1.6 mM CaCl2·2H2O, and 0.4% polyvinylpyrrolidine pH 2.5) (14) in a 35-mm Petri dish. Up to 100 embryos can be processed in the same volume. 3. As soon as the zona pellucida is dissolved, add 1 mL of coculture medium to neutralize the acidic pH. 4. Wash cells in 5–7 mL of fresh aggregation medium (see Note 12).

472

Khillan

Fig. 4. Model for the integration of ES cells into a morula. The bottom of the microwell provides an area for tight cell–cell contact and simulates the formation of ES cell aggregates (A). Transfer of embryos over these cells establishes cell–cell–embryo contact and further growth of the embryo therefore results in integration of ES cells into the embryo (B). 5. Transfer one embryo each to individual microwells over ES cells (Fig. 4) and incubate overnight in the CO2 incubator. 6. Collect blastocyts after overnight culture and wash in 5 mL of aggregation medium. 7. Store embryos in incubator until transfer into pseudopregnant mothers. 8. The animals with more than 90% chimerism can be generated reproducibly by this method and most of the animals display germline transmission.

3.12. Transfer of Embryos into Pseudopregnant Mothers 1. Anesthetize a 2.5-d pseudopregnant female by injecting 0.4–0.5 mL of 2.5% avertin and wipe the back with 70% ethanol. 2. Shave fur on the back and make an incision of about 1 cm in the mid region. Uterine horns on both sides can be reached from the same incision. 3. Place mouse under a dissecting microscope. 4. Cut off connective tissue under the skin very carefully and make an incision in the peritoneal membrane on one side and locate the pink-colored ovary under the membrane. 5. Pull out reproductive organs by pulling the periovarian fat that surrounds the ovary. 6. Mount uterine horn carefully onto a cotton applicator to provide support to the organs. 7. Place Petri dish with embryos under another dissecting microscope and collect 10 embryos in an embryo transfer needle using a minimum volume of medium. A few air bubbles preceding the embryos help to control the flow (see Note 13). 8. With a 25-gage needle, make a hole near the tip of the uterus. Keeping the hole in view, insert the transfer needle into the uterus and, slowly deliver the embryos inside. 9. Remove cotton applicator and with the help of a new applicator, push organs back into the peritoneal cavity. 10. Close the opening with a surgical suture followed by clamping the skin with wound clips. If the cut in the membrane is not very large, the suturing may not be necessary. 11. Place animal under a lamp until recovery. Pups are born after 16–17 d. After 7–10 d the degree of chimerism can be judged by the coat color (Fig. 5).

3.13. Germline Transmission 1. Mate chimeric animals with wild-type females at the age of 7–8 wk (see Note 14). The germline transmission is obvious from the coat color of the newborns. 2. Check for coat color after 7–10 d. Typically, the crossing of a chimeric male with a C57BL6 female will produce a progeny that has an agouti coat color, which represents the contribution of ES cells to the germline. 3. At the age of 3 wk a piece of tail is excised to isolate DNA.

Chimeric Animals and Germline Transmission

473

Fig. 5. Chimeric animals prepared from aggregation of FVB/N embryos (albino) with R1 cells derived from 129SV (agouti) strain of mice. Three chimeras displayed almost complete ES-cell-derived phenotype, whereas the other three were between 70% and 90% chimeric.

Fig. 6. PCR on DNA isolated from the progeny of chimeric animal crossed with a wild-type FVB/N female. The DNA was amplified using primers neo 2 and neo 3, specific for the neomycin gene shown in Fig. 2. The DNA from three pups amplified a band of 330 bp, indicating the germline transmission of the mutant allele. 4. Confirm germline transmission either by polymerase chain reaction PCR or by Southern analysis. 5. Progeny obtained in the first cross are heterozygous for the mutant allele, therefore breeding of two heterozygotes will generate homozygous progeny in which both the alleles are disrupted. (see Note 15).

3.14. Analysis of Animals 1. Cut a small piece of tail of a 2- to 3-wk-old animals and isolate total DNA (13). 2. Mince tail tissue and digest in lysis buffer (50 mM Tris-HCl pH 8.0, 100 mM EDTA, 100 mM NaCl, 1.0% sodium dodecyl sulfate (SDS), and 35 mg/mL proteinase K) for 8–10 h. 3. Extract DNA with phenol:chloroform followed by precipitation with ethanol. 4. Dissolve DNA in TE buffer and analyze either by Southern blot (13) or by PCR analysis. (Figs. 5 and 6, respectively).

474

Khillan

4. Notes 1. Although microinjection of ES cells into blastocysts is a quite efficient method, it has many limitations, such as it requires specialized training and skills and also needs many expensive and sophisticated instruments. The morula aggregation provides an alternate method to prepare the chimeric animals, which does not require any major equipment. 2. The function of PMSG is equivalent to the follicle-stimulating hormone, whereas HCG serves the function of a leutinizing hormone. 3. The ES cells are available as established cell lines from several different laboratories (e.g., R1 cells from Dr. A. Nagy from Canada and J1 from Dr. Rudolph Jaenisch, MIT, MA). The cells are also commercially available (RW4 cells from Genome Systems Inc., St. Louis, MO). The lines are established from the inner cell mass (ICM) of a blastocyst. These cells are pleuripotent and can be maintained in the undifferentiated state indefinitely. When introduced into a host blastocyst, ES cells can contribute to the formation of all organs of the developing embryo, including testis and ovary. The microinjected cells integrate into the ICM of the host blastocyst and contribute to the formation of different tissues. The animal born from such embryo is chimeric (i.e., contains tissue formed from microinjected cells as well as cells from the host blastocyst). The transmission of the gene to the next generation occurs only if the microinjected cells contribute to the germline. The ES cells cultures are carried out on a layer of mitotically inactive fibroblast feeder cells in ES cell culture medium. The pleuripotency of cells can be maintained by adding lymphocyte inhibitory factor (LIF) in the medium which prevents differentiation of ES cells. The fibroblasts feeder cells secrete LIF. In general, the ES cells require about 2000 IU of LIF/mL. The feeder cells provide about 1000 IU and, therefore, the cells are supplemented by exogenous LIF which is commercially available from Gibco-BRL. 4. For the micromanipulation of embryos for aggregation procedure, only one type of micropipets is required, the embryo isolation and transfer needle. The internal diameter of the micropipet is about 100–120 µM. The Glass capillaries for the micropipets are washed overnight in 100% ethanol and rinsed with dH2O followed by heating at 180°C. 5. The shape of the microwell is such that about 15–20 cells settle at the bottom to from an aggregate. The denuded embryo therefore is placed over the cells. 6. Radiolabeled compounds are health hazards; therefore, they must be handled behind a thick protective plastic shield. 7. The feeder layer cells are mitotically inactivated before preparing feeder plates. The cells are treated with mitomycin C, 10 µg/mL, for 2 h. The cells are washed extensively in PF culture medium and can be used directly to prepare feeder plates. Alternatively, the cells are irradiated at 4000–6000 rads and aliquoted as 1 × 106 cells and stored in liquid nitrogen. 8. For the ES cell culture, the medium must be changed every day and the cells must be splitted after 48 h otherwise the cells tend to differentiate. 9. To ensure the availability of 4–5 foster mothers, generally 30 males are sufficient. The day before the collection of foster mothers, approx three CD1 females 5–6 wk of age are mixed with males. The foster mothers are separated everyday by screening the females for vaginal plugs. The females can be allowed to stay with the males during the period the foster mothers are required. However it is recommended to separate females from males over the weekends and mix again when the experiment is to be resumed. Discard females that are more than 12-wk-old and are fat. The efficiency of pregnancy decreases with age. The unused females can be recycled. 10. Female mice are superovulated to obtain synchronized embryos. Usually, 8–10 female mice are sufficient to obtain 30–50 good compact morulas. 11. For the transfer of blastocyts, the foster mothers are generally collected a day later so that the blastocysts are transferred to 2.5-d pseudopregnant females, which is required for the embryo to adjust to the mother’s environment.

Chimeric Animals and Germline Transmission

475

Fig. 7. Total DNA isolated from three independent ES cell clones was analyzed by Southern-blot analysis. The clones were targeted to introduce a null mutation in the gene for collagen type II (COL2A1). The normal allele generates a 9.5 kb band with the COL2A1 specific probe after digestion with BamHI restriction enzyme. A second band of 8.2 kb is produced from the targeted allele. The equal intensity of both the bands represent that one allele is disrupted. Similar pattern was observed when the DNA from the progeny was analyzed. 12. For ES cells to integrate with the embryo, the zone a pellucida must be removed from the embryo. The tyrode solution is highly acidic, therefore it is extremely important that the embryo be exposed for minimum amount of time. The embryos are constantly watched under a steromicroscope during the treatment. It is more helpful to have a pipet full of medium so that as soon as the zona pellucida is dissolved, the tyrode solution is diluted with the fresh medium. After the removal of zona pellucida, the embryos tend to stick to each other, a high concentration of BSA (10 mg/mL) in the aggregation medium prevents adhesion of denuded embryos to each other. 13. Generally, 10 embryos are transferred to each female in 1 uterine horn only. On the average, the efficiency of implantation is about 50%; therefore, five pups in the one uterine horn is an appropriate number. However, the extra embryos can be transferred to the other uterine horn, making a similar incision in the other side of the peritoneal cavity. 14. Almost all of the established ES cell lines are of male origin, therefore the chimeric animals are mostly males. This provides an advantage because males can be mated with several females to obtain positive progeny more efficiently and quickly. 15. If the targeted gene is essential for the development, the homozygous progeny will show either developmental abnormalities or lethal phenotype after birth.

References 1. Bradley, A., Evans, M., Kaufman, M. H. and Robertson, E. (1984) Formation of germline chimeras from embryo-derived teratocarcinoma cell lines. Nature 309, 255,256. 2. Capecchi, M. R. (1989) The new mouse genetics: altering the genome by gene targeting. Trends Genet. 5, 70–76. 3. Evans, M. J. and Kaufman, M. H. (1981) Establishment in culture of pleuripotential cells from mouse embryos. Nature 292, 154–56. 4. Martin, G. R. (1981) Isolation of a pleuripotent cell line from early mouse embryos cultured in medium conditioned by teratocarcinoma stem cells. Proc. Natl. Acad. Sci. USA 78, 7634–7638. 5. Robertson, E. J. (1987) in Teratocarcinomas and Embryonic Stem Cells: A Practical Approach (Robertson, E. J., ed.), IRL, Oxford, UK, pp. 71–112.

476

Khillan

6. Hasty, P., Crist, M., Grompe, M., and Bradley, A. (1994) Efficiency of insertion versus replacement vector targeting varies at different chromosomal loci. Mol. Cell. Biol. 14, 8385–8390. 7. Mansour, S. L., Thomas, K. R., and Capecchi, M. R. (1988) Disruption of the protooncogene int-2 in mouse embryo-derived stem cells: a general strategy for targeting mutations to non-selectable genes. Nature 336, 348–352. 8. Mortensen, R. (1993) In Current Protocols in Molecular Biology, Vol. 1. (Susubel, F. M., et al., eds.), Wiley, New York, pp. 9. 15. 1–9. 17. 3. 9. Wood, S. A., Pascoe, W. S., Schmidt, C., Kemler, R., Evans, M. J., and Allen, N. D. (1993) Simple and efficient production of embryonic stem cell-embryo chimeras by coculture. Proc. Natl. Acad. Sci. USA 90, 4582–4585. 10. Nagy, A., Gocza, E., Diaz ,E. M., Prideaux, V. R., Ivanyi, E., Markkula, M., and Rossant, J. (1990) Embryonic stem cells alone are able to support fetal development in the mouse. Development 110, 815–821. 11. Nagy, A., Rossant, J., Nagy, R., Abramow-Newerly, W., and Roder, J. C. (1993) Derivation of completely cell culture-derived mice from early-passage embryonic stem cells. Proc. Natl. Acad. Sci. USA 90, 8424–8428. 12. Khillan J. S. and Bao, Y. (1997) Preparation of animals with a high degree of chimerism by one-step coculture of embryonic stem cells and preimplantation embryos. BioTech 22, 544–549. 13. Hogan, B., Beddington, R., Costantini, F., and Lacy, E. (1994) Manipulating the Mouse Embryo. Cold Spring Harbor Labs., Plainview, NY.

Gene Knockout Using Cre Recombinase

477

45 Conditional Gene Knockout Using Cre Recombinase Yunzheng Le and Brian Sauer 1. Introduction The directed introduction of null mutations into defined genes has proven invaluable in elucidating gene function in a variety of experimental organisms. In the last decade or so this approach has been extended to mice (1) by the combined use of homologous recombination in murine embryonic stem (ES) cells to precisely target a mutation to a desired gene and subsequent derivation of mice carrying the targeted gene alteration from the genetically manipulated ES cells (e.g., by injection of genemodified ES cells into blastocysts with subsequent germline transmission). In most instances null, or knockout (KO), mutations have been generated in mice by either simple insertion of a neo selectable marker in the target gene or neo insertion coupled with deletion of a critical region of the target gene. Targeted null mutations in a gene of interest, however, can lead to embryonic lethality in mice, thus obscuring the particular role of that gene in a target tissue or in the adult. Site-specific recombination strategies allow circumvention of the problem of embryonic lethality by directing gene ablation in a spatially and temporally controlled manner. Because the Cre recombinase of phage P1 catalyzes efficient excisive recombination in mammalian cells, it has been become a useful tool for generating a conditional KO (2,3). In addition, Cre-mediated excision has been useful both for targeted activation of genes in transgenic mice and for elimination of the selectable drug marker that is necessarily left in the genome after homologous gene targeting in ES cells (3,4) (see Note 1). The 38 kDa Cre recombinase catalyzes DNA recombination between specific 34-bp sequences called loxP (5). This site exhibits two inverted 13-bp repeats and a central asymmetric 8-bp core region that confers an overall directionality to the site (Fig. 1, see Note 2). Cre-mediated recombination between two directly repeated loxP sites on a DNA molecule results in excision of DNA between the two loxP sites. Mutant sites, such as loxC2, having base changes in the outer 4 bp of one of the inverted repeats (positions 1–4 or 10–13) are also recognized by Cre recombinase (6). The loxC2 site is a useful alternative to loxP when the site will be placed into transcribed sequences because it decreases stability of the 13-bp inverted repeat-derived hairpin in RNA. Such hairpins in the 5' untranslated RNA region can diminish translation. From: Methods in Molecular Biology, Vol. 136: Developmental Biology Protocols, Vol. II Edited by: R. S. Tuan and C. W. Lo © Humana Press Inc., Totowa, NJ

477

478

Le and Sauer

Fig. 1. Recombination sites for Cre recombinase. The wildtype loxP and variant loxC2 sites are shown, with base pair (bp) positions indicated below. The 13-bp inverted repeats are indicated by the thin arrows above the sequence. The loxC2 site differs from the wt site by replacing the T at position 2 with a C, and is named accordingly. By convention, the sense of the overall directionality of the lox site is as indicated by the large arrow. Note that loxG33 would be functionally equivalent to loxC2, because of symmetry.

1.1. Strategies for Genomic Manipulation Conditional gene deletion allows assessment of a gene’s function in a target tissue without disturbing expression of that gene in nontarget tissues. Two components are required: a target mouse carrying a gene-modified allele of the gene to be ablated, and a cre transgenic mouse that expresses Cre under the control of a promoter with the desired spatial and temporal pattern of expression. Exact placement of lox sites in the target gene will depend both on the type of deletion event desired and on constraints imposed by the structure of the target gene. An example of a target-gene modification strategy in ES cells for the generation of a conditional knockout is shown in Fig. 2A (see Note 3). The neo selectable marker, flanked by two directly repeated lox sites (a lox2 neo cassette), is placed at one of the deletion endpoints (shown here in the first intron) and a third lox site is placed at the other deletion endpoint (shown here in the 5' leader before the first ATG in exon I) (see Note 4). Standard homologous recombination in ES cells is used to modify the target-gene locus (shown here by a double crossover event). Because neo could interfere with the correct expression of the target gene (depending on its placement in a particular strategy), it may be prudent to remove it by a limited Cre-mediated recombination event that removes the neo interval but leaves intact the genomic interval that is to be later targeted for deletion. Removal of neo can be effected in ES cells by transient transfection with a Cre-expressing construct. As shown in Fig. 2B, conditional gene KO is accomplished by the mating of two animals, one having the lox-modified target locus (most conveniently a homozygote) and the other carrying the cre transgene and, in addition, one copy of either a null allele in the target locus or a copy of the lox-modified target locus. Cre-mediated recombination then excises the target gene from the genome in a manner that reflects the (tissuespecific) pattern of expression of the cre gene. Penetrance of expression of the cre transgene will govern the ratio of deletion vs nondeletion on a per cell basis in the target tissue (see Note 3). A similar mating strategy is used for targeted activation of a transgene or endogenous gene in a tissue-specific, developmental or inducible fashion (Fig. 3). A “STOP”

Gene Knockout Using Cre Recombinase

479

Fig. 2. A conditional KO strategy. (A) Targeting of the genomic locus by homologous recombination generates a modified gene that carries the correctly placed lox sites (indicated by white arrows) and the neo selectable marker (black rectangle). A second transfection step with a cre expression plasmid can be used to give partial excision. ES cells with the desired genomic structure are injected into blastocysts to give chimeric founders. (B) Mating strategy for a conditional KO. The desired double “transgenic” inherits a conditional allele from one parent and the cre transgene and the second copy of the conditional allele from the other. Note that the cre transgenic parent used in this cross is the product of a previous cross of a cre transgenic with the conditional knockout mouse so that it is both heterozygous at the target knockout locus and positive for the cre transgene. Tissue-specific Cre expression is shown here by the black ears. In the Cre+ mouse homozygous for the conditional allele, productive recombination in the ear is represented by the white ears.

cassette flanked by two directly repeated loxP sites (a lox2 STOP cassette) is placed between the promoter and the gene to be activated (4). STOP is designed to block gene

480

Le and Sauer

Fig. 3. Recombinational activation of gene expression. A synthetic STOP sequence (GenBank Accession No. U51223), is flanked by two directly repeated lox sites (white arrows) and is placed between a promoter and the target gene to be activated. Cre expression removes the STOP signal to allow target gene expression, leaving behind a single 34-bp lox site.

expression and consists of a stuffer region (from the yeast HIS3 gene), the SV40 polyadenylation region, and errant optimized ATG translational start and splice donor signals. Cre-mediated excision removes STOP thus permitting target-gene expression under the control of the adjacent promoter. Two animals are required: one expressing Cre with the desired spatial and temporal pattern and the other either carrying a lox2 STOP-equipped transgene or, alternatively, an animal modified by homologous genetargeting in ES cells to carry the lox2 STOP cassette in the desired endogenous gene. Mating of these two animals results in activation of the target gene in those cells that both express (or have expressed in a progenitor cell) Cre recombinase and are capable of expressing the target gene, as determined by the specificity of the promoter used in the lox target strain. To target recombination to a particular tissue, Cre should be specifically expressed in the target tissue. This can be achieved by making a transgenic mouse with cre under the control of a promoter with the desired expression characteristics. Alternatively, the cre gene can be targeted to be under the control of an endogenous promoter in the genome by using homologous recombination in ES cells (a “knock-in”). The pattern of Cre mouse expression should be checked using a Cre-specific antibody (2,7), rtPCR, or in situ RNA hybridization. This is important because some promoters are susceptible

Gene Knockout Using Cre Recombinase

481

to position effects, and resulting founder lines may express the transgene in a mosaic fashion in the target tissue or misexpress the transgene in an unwanted tissue. Gene ablation (or gene activation) will be difficult to achieve in the majority of cells in a target tissue if Cre is expressed in a variable and mosaic fashion. cre knock-ins may thus be preferred unless the promoter used in a transgenic strategy has already been established to behave with the desired expression pattern in a transgenic (e.g., by monitoring expression of a lacZ reporter gene). 2. Materials 2.1. Detection of the cre Transgene by PCR 1. DNA oligonucleotide primers are stored at –20°C at a stock concentration of 3 µM. Sense: 5' AGG TGT AGA GAA GGC ACT TAG C 3' Antisense: 5' CTA ATC GCC ATC TTC CAG CAG G 3'. 2. 10X PCR buffer: 500 mM KCl, 100 mM Tris-HCl pH 8.3, 20 mM MgCl2, stored at –20°C. 3. dNTP’s: 12.5X stock solution, 2.5 mM of each in H2O, stored at –20°C. 4. AmpliTaq polymersase (Perkin-Elmer, Norwalk, CT) at 250 U/50 µL, stored at –20°C. 5. Perkin Elmer 9600 thermocycler, or equivalent, and 0.2 mL PCR reaction tubes.

2.2. Antibody Detection of Cre Protein in Cells 1. 2. 3. 4. 5.

Anti-Cre mAb 7.23 (Berkeley Antibody Co., Richmond, CA), 1 mg/mL. Formaldehyde (Electron Microscopy Sciences, Ft. Washington, PA), 16% solution. Normal goat serum (Life Technologies, Inc., Gaitherburg, MD). FITC-conjugated goat anti-mouse IgG1 polyclonal serum (Southern Biotech., Birmingham, AL). Phosphate buffered saline (PBS); PBS/B (PBS + 0.5% BSA + 0.01% NaN3); PBS/B + S (PBS/B + 0.5% saponin). Filter with 0.2-micron filter. 6. Epifluorescence microscope for FITC detection.

2.3. Gene Popout in ES Cells by GFPcre All cell culture reagents are from Life Technologies, Inc. unless otherwise indicated. 1. 2. 3. 4. 5.

6. 7. 8. 9. 10.

Culture dishes: Falcon 3002 (6-cm). Phosphate buffered saline without calcium and magnesium. 0.2% Gelatin (Sigma) in PBS, autoclaved. Irradiated mouse embryonic fibroblasts (EFB, made in the laboratory). ES medium: DMEM suplemented with 20% fetal calf serum (FCS, Hyclone, Logan, UT), 1000 U/mL ESGRO, 2 mM L-glutamine, 0.1 mM of nonessential amino acids and 0.1 mM of β-mecaptoethanol, with 50 U/mL of penicillin, and 50 µg/mL of streptomycin. 0.1% trypsin in PBS with 0.05 mM EDTA. Plasmid pBS500, EF1a-GFPcre (8). 2.5 M CaCl2. 50 mM BES solution: 50 mM, N,N-bis[2-hydroxyethyl]-2-aminoethanesulfonic acid, 280 mM NaCl, and 1.5 mM Na2HPO4, adjust pH to 6.96 with HCl. Epifluorescence microscope for FITC (GFP) detection.

3. Methods 3.1. Detection of the cre Transgene by PCR After identification of cre transgenic animals by Southern blotting, it is convenient to monitor cre transgene transmission by PCR, using standard genomic “tail DNA” (9).

482

Le and Sauer

PCR amplification yields a 411 bp product diagnostic for cre (GenBank Accession No. X03453). 1. In a 0.2-mL reaction tube add: 50–500 ng of genomic DNA, 5 µL of 10X PCR buffer, 4 µL of 10X dNTPs, 5 µL of each sense, and antisense primers, and H2O to 50 µL total volume. 2. Denature 5 min at 94°C. 3. Add 0.5 µL of AmpliTaq polymerase. 4. Perform 25 or 30 cycles of amplification: 30 s at 94°C, 30 s at 63°C, 1 min at 72°C.

3.2. Antibody Detection of Cre Protein in Cells Indirect immunofluorescence with anti-Cre specific antibody is used to assess tissue specific expression and also the level of mosaicism of expression in a population of cells. Detection by antibody quickly identifies Cre-expressing animals that may be suitable for further analysis (7). 1. Cre-expressing cells are washed once with PBS and fixed with PBS + 2% formaldehyde for 20 min at RT. All steps are at RT unless otherwise indicated. 2. Wash twice with PBS and once with PBS/B. 3. Permeabilize with PBS/B + S for 6 min. 4. Block with PBS/B + 5% normal goat serum for 30 min. 5. Incubate with anti-Cre mAb 7.23 (diluted 1/100 in PBS/B + S) for 15 min. Incubate longer if necessary. 6. Wash three times with PBS/B + S. 7. Incubate with FITC-conjugated goat antimouse IgG1 polyclonal serum (10 µg/mL in PBS/ B + S containing 7% normal goat serum) for 20 min. 8. Wash three times with PBS/B + S. 9. Examine by epifluorescent microscopy.

3.3. Gene Popout in ES cells by GFPcre After targeted modification of a locus in ES cells by homologous recombination, removal of the selectable marker (for example neo) may be desired in order to preclude interference by the marker gene on correct gene expression at the target locus. Incorporation of a lox2 neo cassette into the homologous targeting vector allows subsequent Cre-mediated removal of the neo gene from the targeted locus. This step is facilitated by use of a functional fusion between Cre and an enhanced fluorescent derivative of the green fluorescent protein (GFP) of Aequorea victoria (10). Transient transfection with the GFPcre fusion construct pBS500 (8) results in cells that are simultaneously GFP+ and Cre+ (and, hence, committed to excision of neo). Because the transfection efficiency of ES cells is often low, it is convenient to enrich for productively transfected ES cells by using fluorescence (Fig. 4). We have found that gene transfer by calcium phosphate co-precipitation with DNA (11) to give somewhat more routinely efficient transient transfection than electroporation. Alternatively, standard electroporation of ES cells (12) with pBS500 (EF1a-GFPcre) can be used. 1. Plate 2 × 106 ES cells on a gelatinized 6-cm dish seeded with 2 × 106 irradiated EFB two days before the transfection. Change medium next day and also on the second day, 3 h before transfection. 2. Combine 9 µg pBS500 DNA with H2O (total volume for DNA and H2O is 165 µL), add 18.3 µL 2.5 M CaCl2 and mix. Add the DNA solution dropwise into 183 µl BES solution, mixing gently and continuously. Incubate 20 min at RT.

Gene Knockout Using Cre Recombinase

483

Fig. 4. GFPCre expression in ES cells. ES cells were transiently transfected with pBS500 (EF1α-GFPcre) by CaPO4 coprecipitation. Two days later cells were trypsinized and examined for fluorescence. 3. Wash the dish of ES cells with 5 mL PBS, add 2 mL 0.1% trypsin solution and incubate 2 min at RT. Remove the trypsin solution and incubate the dish at 37°C for 5 min. Add 4 mL medium to inactivate the trypsin. Count the cells and dilute the ES cells in the medium to 3 × 105 cell/mL. This ES cell suspension will be mixed with the DNA/CaPO4 coprecipitate. 4. Mix the DNA/CaPO4 coprecipitate solution with 5 mL of the ES cell suspension (1.5 × 106 ES cells total). Plate the mixture on a 6-cm gelatinized dish and incubate overnight at 37°C. 5. Wash the transfected ES cells with DMEM and replace with fresh medium 16 h after transfection. 6. Fluorescence should be observed in about 2–7% of the GFPcre-transfected cells 2 d after transfection. After trypsinization, fluorescent (Cre+) ES cells can be either sorted manually at this time with a micromanipulator, or by FACS.

4. Notes 1. Genomic manipulation strategies with the distantly-related DNA recombinase FLP (from yeast 2-µ circle) are conceptually similar to those for Cre recombinase. Somewhat greater variability in efficiency of recombination in mammalian cells has been reported for FLP (13,14). 2. There are two “ATG’s” in the conventional orientation of loxP as shown in Fig. 1. However, in the reverse orientation there are no ATGs, hence the reversed orientation is preferred when inserting the lox sequence into the 5' RNA leader region of a gene. 3. In the conditional KO strategy, the cre transgenic parent in the “knockout activation” cross is heterozygous at the target KO locus for the conditional allele. Alternatively, the cre parent can be heterozyous at the target locus for a wt and a complete null. In this case, it should be verified that animals heterozygous for the null allele at this locus are phenotypically wt. 4. To evaluate the recombination potential of plasmids containing two loxP sites, it is convenient to transform the plasmid into the Cre-expressing E. coli strain BS591 [F– recA1 endA1 hsdR17 δlac(lacZYA-argF)U169 supE44 thi-1 gyrA96 (λ imm434 nin5 X1-cre)].

484

5. 6. 7.

8.

9. 10.

Le and Sauer Plasmid DNA isolated from this strain will have undergone Cre-mediated recombination and can be checked by a simple restriction digest (15). Alternatively, and more rigorously, lox regions should be sequenced in the final plasmid construction. PBS + 0.1% Triton X-100 for 6 min also gives permeabilization of cells for indirect immunofluorescent detection of Cre. Cre protein can be detected in target tissues by standard Western blotting or immunoprecipitation by using the anti-Cre mAb 7.23 at dilutions of 1/1000 and 1/150, respectively. The pattern of Cre expression from a cre transgenic or knock-in animal may also be determined using the GFPcre gene. This potentially would allow direct detection of Cre but most likely would only be useful in situations in which the promoter used is known to give relatively strong expression (to allow direct detection of fluorescence) in the target tissue. Since only a single catalytic event is required for productive recombination in a cell, the actual amount of Cre required per cell is not high, and antibody detection is potentially more sensitive than reliance on fluorescence from GFPCre. As an alternative to excision of the loxP-flanked neo gene in ES cells by transient transfection with a cre plasmid, ES cells can be injected directly into blastocysts to give a chimeric founder that is then mated with the EIIa-cre mouse (16). The EIIa promoter directs Cre expression only in fertilized zygotes and early embryos, and is not expressed post-implantation, thus allowing neo removal from the germline. In multi-lox transgene arrays, both partial and complete deletion events can be obtained that are then transmitted through the germline. GFPCre localizes to the cell nucleus (8) since Cre carries an endogenous nuclear localization signal (unpublished). Following sorting of ES cells transiently transfected with GFPcre, resulting colonies should be confirmed for excisive recombination by Southern or a suitable PCR assay. Most (≥80%) fluorescent cells will give rise to colonies carrying the desired lox-delimited deletion. DNA analysis will also confirm that the colony is not mosaic for both recombinant and nonrecombinant cells.

References 1. Gordon, J. W., Harold, G., and Leila, Y. (1993) Transgenic animal methodologies and their applications. Hum. Cell 6, 161–169. 2. Sauer, B., and Henderson, N. (1988) Site-specific DNA recombination in mammalian cells by the Cre recombinase of bacteriophage P1. Proc. Natl. Acad. Sci. USA 85, 5166–5170. 3. Gu, H., Marth, J. D., Orban, P. C., Mossmann, H., and Rajewsky, K. (1994) Deletion of a polymerase beta gene segment in T cells using cell type-specific gene targeting. Science 265, 103–106. 4. Lakso, M., Sauer, B., Mosinger, J., B., Lee, E. J., Manning, R. W., Yu, S.-H., Mulder, K. L., and Westphal, H. (1992) Targeted oncogene activation by site-specific recombination in transgenic mice. Proc. Natl. Acad. Sci. USA 89, 6232–6236. 5. Hoess, R. H., and Abremski, K. (1990) The Cre-lox recombination system, in, Nucleic Acids and Molecular Biology, vol. 4. (Eckstein, F. and Lilley, D. M. J., eds.), SpringerVerlag, Berlin, pp. 99–109. 6. Sauer, B., Whealy, M., Robbins, A., and Enquist, L. (1987) Site-specific insertion of DNA into a pseudorabies virus vector. Proc. Natl. Acad. Sci. USA 84, 9108–9112. 7. Schwenk, F., Sauer, B., Kukoc, N., Hoess, R., Müller, W., Kocks, C., Kühn, R., and Rajewsky, K. (1997) Generation of Cre recombinase-specific monoclonal antibodies to characterize the pattern of Cre expression in cre-transgenic mouse strains. J. Immunol. Meth. 207, 203–212.

Gene Knockout Using Cre Recombinase

485

8. Gagneten, S., Le, Y., Miller, J., and Sauer, B. (1997) Brief expression of a GFPcre fusion gene in embryonic stem cells allows rapid retrieval of site-specific genomic deletions. Nucleic Acids Res. 25, 3326–3331. 9. Gendron-Maguire, M., and Gridley, T. (1993) Identification of transgenic mice. Meth. Enzymol. 225, 794–799. 10. Heim, R., Cubitt, A. B., and Tsien, R. Y. (1995) Improved green flourescence. Nature 373, 663–664. 11. Chen, C., and Okayama, H. (1987) High-efficiency transformation of mammalian cells by plasmid DNA. Mol. Cell. Biol. 7, 2745–2752. 12. Ramírez-Solis, R., Davis, A. C., and Bradley, A. (1993) Gene targeting in embryonic stem cells. Meth. Enzymol. 225, 855–878. 13. Feiring, S., Kim, C. G., Epner, E. M., and Groudine, M. (1993) An “in-out” strategy using gene targeting and FLP recombinase for the functional dissection of complex DNA regulatory elements: analysis of the β-globin locus control region. Proc. Natl. Acad. Sci. USA 90, 8469–8473. 14. O’Gorman, S., Fox, D. T., and Wahl, G. M. (1991) Recombinase-mediated gene activation and site-specific integration in mammalian cells. Science 251, 1351–1355. 15. Sauer, B., and Henderson, N. (1988) The cyclization of linear DNA in Escherichia coli by site-specific recombination. Gene 70, 331–341. 16. Lakso, M., Pichel, J. G., Gorman, J. R., Sauer, B., Okamoto, Y., Lee, E., Alt, F. W., and Westphal, H. (1996) Efficient in vivo manipulation of mouse genomic sequences at the zygote stage. Proc. Natl Acad. Sci. USA 93, 5860–5865.

Application of Cre/loxP in Drosophila

487

46 Application of Cre/loxP in Drosophila Site-Specific Recombination and Transgene Coplacement Mark L. Siegal and Daniel L. Hartl 1. Introduction The use of site-specific recombinases has revolutionized the genetic analysis of development and has made possible the precise engineering of genomes (1,2). In Drosophila, the FLP/FRT system, introduced by Golic and Lindquist (3), has been used (1) to generate genetic mosaics by mitotic recombination as well as by “flip-outs” (3–5), and (2) to generate defined chromosomal rearrangements (6,7). In yeast and mammalian cells, site-specific recombination has also been used to mediate targeting of exogenous DNA to genomic docking sites (8). Although such targeted integration is by nature an inefficient process—as a result of the favoring of intramolecular over intermolecular recombination—this limitation has been overcome in these systems by the ability to introduce DNA into a large number of cells simultaneously and to select for rare integration events by chemical means. In Drosophila, such an approach is not currently available, although an approximation of targeted integration of exogenous DNA has been used, with varying efficiency, to mobilize FRT-flanked DNA already in the genome to a specific FRT target site elsewhere (9). An efficient, reliable targeted-integration system in Drosophila would reap enormous practical benefits. Because transgenes are subject to genomic position effects, independent integrations of the same transformation construct may differ quantitatively and qualitatively in their expression (10–13). Thus, the ability to compare transgenes at the same position in the genome would increase the power of an experiment to detect subtle expression differences between the transgenes. Furthermore, some experiments absolutely require that transgenes be present at the same position, such as those investigating pairing or location-dependent phenomena—including transvection (14), meiotic recombination (15), and position-effect variegation (16)—and those seeking to trace allele frequencies over many generations in experimental populations so as to make inferences concerning the relative effects of the transgenes on reproductive fitness (17). In this chapter, we outline the use of a system we have developed to place pairs of transgenes at the same position in the Drosophila genome (18). This system of From: Methods in Molecular Biology, Vol. 136: Developmental Biology Protocols, Vol. II Edited by: R. S. Tuan and C. W. Lo © Humana Press Inc., Totowa, NJ

487

488

Siegal and Hartl

transgene coplacement relies only on the efficient procedures of P-element-mediated germline transformation and the excision reactions of the FLP and Cre site-specific recombinases. The pair of transgenes is introduced on a single, specially-designed P element so that the action of FLP eliminates one transgene, whereas the action of Cre eliminates the other transgene. Thus, one may derive, from each insertion of the original P element, pairs of strains containing one or the other of the transgenes, in the same genomic position and orientation, and flanked by the same sequences. In addition to an outline of the coplacement method, we present, here, detailed information on the Cre constructs we have introduced into Drosophila. Because it has the same mode of action as FLP, Cre may be used in all applications for which FLP is appropriate. More importantly, as this chapter demonstrates, the use of Cre and FLP in concert can perform functions that neither one alone can. 2. Materials 2.1. Plasmids 1. The plasmid pP[wFl], shown in Fig. 1A, is a P-element transformation vector that contains two unique cloning regions in which to place each of the transgenes to be compared. In addition, the vector contains the mini-white selectable marker (19), positioned such that it is excised along with the desired transgene upon FLP or Cre recombination. Thus, each insertion of the original transgene-containing P[wFl] construct is identified by the white+ phenotype it confers to otherwise w– flies, whereas desired excisions are identified by reversion of the phenotype to white–. 2. The plasmid pP[SFl], shown in Fig. 1B, is a derivative of pP[wFl] in which the mini-white marker has been removed, leaving a unique SphI restriction site in its place. This vector is useful in cases where a selectable marker other than mini-white is desired, or where the transgenes themselves confer a selectable phenotype.

2.2. Fly Stocks Expressing Cre 1. The Cre-expressing transgene used for transgene coplacement and other applications in Drosophila is a chimeric construct, described in ref. 18, in which the Cre coding sequence is flanked by sequences derived from the active mariner transposable element Mos1 (20), which provides a promoter sequence as well as a translation initiation consensus sequence (21) and a polyadenylation signal sequence. Additionally, an hsp70 promoter lies upstream of the Mos1 promoter. The chimeric Cre construct was cloned into both the mini-whitecontaining P-element vector, pCaSpeR4 (19), and the mini-yellow-containing P-element vector, pCar-y (22), to produce P[w+, cre] and P[y+, cre], respectively. Despite carrying the same Cre construct, these Cre sources have different properties (see Note 1). 2. Because the Cre sources described above are highly expressed without heat shock (see Table 1 and Note 2; ref. 18), we mutagenized flies carrying a second-chromosome insertion of P[y+, cre] to generate Cre sources with inducible expression. The resulting strains (see below and Note 2) are designated P[y+, cre*]. 3. Available stocks include the following (genotypes not specifically described here are in ref. 23; P[ry+, hsFLP] is described in ref. 3): i. insertions of P[w+, cre] on each major autosome ii. insertions of P[y+, cre] on each major chromosome iii. y w, P[y+, cre]; CyO/Sco iv. y w, P[y+, cre]; TM3 Sb/D v. y w; CyO, P[w+, cre]/Sco

Application of Cre/loxP in Drosophila

489

Fig. 1. Transgene coplacement plasmid vectors: (A) The 9.3-kb vector, pP[wFl], showing P-element ends (shaded boxes), loxP sites (open arrowheads), FRT sites (solid arrows), key restriction sites, and the two unique cloning regions (note that the HindIII site in region 2 is not unique). Numbering starts with the HindIII site upstream of mini-white, as it does for the pCaSpeR4 vector from which it is derived (GenBank accession number for pCaSpeR4 is X81645). (B) The 5.3-kb vector, pP[SFl], showing the same features as for pP[wFl]. In addition, a third unique cloning region, “M”, is indicated, into which may be placed a marker of choice (note that the BamHI site in region 1 is no longer unique). (C) Footprint of recombination of either P[wFl] or P[SFl] by either Cre or FLP. Recombination by Cre leaves the transgene cloned into region 2, flanked on one side by a loxP site and on the other side by an FRT site; recombination by FLP leaves the transgene cloned into region 1, flanked by the same sequences. Details of construction of pP[wFl] and pP[SFl] are available on our World Wide Web site at http://www.oeb.harvard.edu/hartl/lab/, or upon request from the authors. vi. y w; P[y+, cre*] vii. y w; CyO, P[y+, cre*]/Sco (remobilization of P[y+, cre*]) viii. y w; TM6B, P[w+, cre]/MKRS, P[ry+, hsFLP] 2.3. All molecular techniques, including cloning, Southern hybridization, and fluorescent in situ hybridization to polytene chromosomes are performed using standard reagents and protocols (24,25). 2.4. Flies are raised on standard cornmeal-molasses medium and at 25°C unless otherwise noted.

3. Methods 1. Cloning of transgene coplacement constructs: Introduce into the unique cloning sites of either pP[wFl] or pP[SFl] the pair of transgenes to be coplaced (see Fig. 1 and Notes 3 and 4). The transgene introduced into cloning site 1 will be excised by Cre, and the transgene introduced into cloning site 2 will be excised by FLP.

490

Siegal and Hartl Table 1 Germline Excision Frequencies of Various Cre and loxP Combinations loxP target P[^a>w+^m>] h.s. no h.s. h.s.

P[^w+^] Cre source

no h.s.

y w, P[y+, cre]; Cyo/Sco

160/160 (100.0) 345/345 (100.0) 298/298 (100.0) 96/277 (34.7) 175/175 (100.0)

y w; CyO, P[w+, cre]/Sco yw; P[y+, cre] a y w; CyO, P[y+, cre*]/Sco y w; TM6B, P[w+, cre]/ MKRS, P[ry+, hsFLP]

98/98 (100.0) † 258/258 (100.0) 344/388 (88.7) †

76/106 (71.7) 88/117 (75.2) 171/233 (73.4) 0/83 (0.0) 156/315 (49.5)

95/104 (91.3) 127/130 (97.7) 149/157 (94.9) 15/270 (5.6) 151/152 (99.3)

Note: Two loxP target transgenes, described in (18), were used to assay Cre activity. P[^w+^] contains a mini-white gene flanked by loxP sites, whereas P[^a>w+^m>] is a derivative of P[wFl] in which the Adh genes of D. affinidisjuncta (“a”) and D. melanogaster (“m”) are cloned into cloning sites 1 and 2 (for an explanation of notation, see Note 8). Each Cre stock was crossed separately with both P[^w+^] and P[^a>w+^m>] stocks, and these crosses were brooded once. The first brood was heat shocked (h.s.) as first-instar larvae for 1 h at 37°C, and the second brood was not heat shocked (no h.s.). Appropriate progeny genotypes were backcrossed to y w flies to assay germline excision frequencies, reported as the fraction of progeny in which the loxP target was excised (percentages are in parentheses). Those crosses exhibiting lethality upon heat shock (see Note 2) are marked by †. aThis is the second-chromosome insertion of P[y+, cre] that was mutagenized to produce P[y+, cre*].

2. Transformation of constructs: Use standard means (26,27) to generate P-element-mediated germline transformants of the transgene coplacement constructs. By genetic segregation and molecular hybridization analyses, identify single-insertion lines that carry the construct on known chromosomes. 3. Selective elimination of transgenes: Pass the original transformed constructs through flies that express either FLP or Cre. If pP[wFl] is used, loss of white+ phenotype in the progeny of such flies will indicate excision of the desired sequences. Induction of FLP expression requires heat shock; generally it is sufficient to heat shock the first-instar larvae in glass vials for 1 h in a circulating water bath at 37°C (3). Depending on transgene size, heat shock for Cre expression may or may not be necessary (see Table 1 and Note 2). 4. Creation of fly stocks to assay: Generate homozygous or balanced lines that carry the FLPor Cre-mediated excision derivatives. For many applications it will be necessary to create lines that are also homozygous for mutations in the gene of interest. An example of such a crossing scheme involving coplaced Alcohol dehydrogenase (Adh) genes is shown in Fig. 2. 5. Confirmation of site-specific recombination: By molecular means, confirm that the desired recombination events have taken place. See Fig. 8 in Siegal and Hartl (18) for an example of fluorescent in situ hybridization analysis of coplaced Adh genes. As shown in Fig. 3, Southern hybridization may be used to confirm the identity of the unexcised transgene and to confirm that flanking genomic sequences are the same for each pair of excision

Application of Cre/loxP in Drosophila

491

Fig. 2. Crossing scheme for coplaced Adh genes of D. affinidisjuncta (“a”) and D. melanogaster (“m”): A third-chromosome insertion of the Adh-containing pP[wFl] derivative, P[^a>w+^m>], is subjected to Cre or to FLP by crossing with males of the strain y w; MKRS, P[ry+, hsFLP]/TM6B, P[w+, cre] (here the hsFLP and cre constructs are abbreviated “FLP” and “cre”; for an explanation of the “>“ and “^” notation, see Note 8). Males are used so as to avoid maternal-effect Cre expression in FLP progeny (see Note 1). Excision derivatives are identified in the next generation by loss of w+. These derivatives are made homozygous on the third chromosome while the second chromosome is made homozygous for a null allele of Adh. Note that the final stocks have a standardized genetic background, as all major chromosomes are from known sources: X, Y, and second chromosomes from Adhn cn; TM3 Sb/D and third chromosomes from the y w stock originally used for transformation of P[^a>w+^m>]. derivatives. PCR analysis using primers specific to each transgene may also be a useful means of characterizing excision derivatives. 6. Gene-expression assays: Assay gene expression in the pairs of lines, by appropriate means. Choice of sample size (number of pairs of lines and number of assayed individuals per line) depends on the variability associated with the measurement across genomic locations and on the correlation in expression between transgenes located at the same genomic position (see Note 6). Our results for ADH specific activity suggest that, in general, this correlation is indeed very high (data not shown).

4. Notes 1. There is a maternal effect associated with the Cre-expressing transgenes, P[w+, cre] and P[y+, cre], in that cre– progeny of cre-heterozygous mothers exhibit recombination at loxP

492

Siegal and Hartl

Fig. 3. Confirmation of site-specific recombination: The FLP-mediated excision derivative of P[^a>w+^m>], carrying the Adh gene of D. affinidisjuncta, is shown schematically on the right, above the corresponding Cre-mediated excision derivative, carrying the D. melanogaster Adh gene. Symbols are as in Fig. 1, and EcoRI sites (RI) are indicated. The extent of the probe sequence used for the Southern hybridizations shown at left is indicated by the striped bars. Genomic DNA was isolated from seven pairs of strains representing FLP or Cre derivatives of third-chromosome insertions of P[^a>w+^m>]. DNA was digested with EcoRI, separated on an agarose gel, blotted onto a nylon filter, and probed with the indicated sequence. Those strains expected to be left with the D. affinidisjuncta Adh are shown in the top autoradiograph, and those expected to be left with the D. melanogaster Adh are shown in the bottom autoradiograph; strains in both blots are ordered the same, according to the original P[^a>w+^m>] insertion strain from which they derive. The Southern blots confirm the identity of each Adh gene (constant bands across all lanes, noted by arrows), as well as the presence of the same 3'flanking genomic sequences for each pair of strains (corresponding variable bands for each lane in top and bottom autoradiographs).

targets (18). The maternal effect is much stronger for P[w+, cre] than for P[y+, cre]. While this difference should not affect choice of Cre source when the cre transgene is present in the same genotype as the loxP target, there may be certain crossing schemes in which the maternal effect may be useful; in such cases, P[w+, cre] should be used. 2. As can be seen in Table 1, the apparent degree of heat-shock inducibility of the Cre sources is dependent on the amount of DNA intervening between loxP sites in the target. For small targets (e.g., P[^w+^]), the excision reactions of P[w+, cre] and P[y+, cre] are practically 100% efficient without heat shock. In fact, for unknown reasons, heat shock is lethal when P[^w+^] is present in the same genotype as P[w+, cre]. This may suggest that the Drosophila genome contains cryptic loxP sites, which may participate in illegitimate recombination events when Cre concentration is extremely high. Excision efficiency without heat shock drops off for larger targets (e.g., P[^a>w+^m>]), but heat shock restores efficiency to high levels, generally without a significant effect on viability. The P[y+, cre*] strain has lower baseline Cre activity in the germline (34.7% excision of P[^w+^]), but high activity upon heat shock (88.7% excision of P[^w+^]). This difference is even more pronounced in the soma; heat-shocked flies carrying both P[y+, cre*] and P[^w+^] have predominantly white eyes with small patches of color, whereas in the absence of heat shock such flies

Application of Cre/loxP in Drosophila

3.

4. 5.

6.

7.

8.

493

have predominantly white+ eyes with small patches lacking color (not shown). Therefore, this mutant recombinase may be particularly useful for application of Cre/loxP to mosaic analysis. Depending on the size of the genes to be coplaced, the resulting plasmid construct may be rather large, and care should be taken both in its construction and in large-scale isolation of pure DNA for embryo injections. In our hands, the construction of large plasmids is often facilitated by using a three-piece ligation strategy (28), which obviates the need to handle large restriction fragments and also yields a low background of undesired clones. Additionally, some P-element-containing plasmids are prone to rearrangements when grown in large bacterial cultures. In our hands, pooling small cultures (2 mL each) after overnight growth is a means of avoiding this problem. Despite these caveats, cloning and transformation of large constructs is routine in most laboratories. For our 20-kb Adhcontaining P element (23-kb plasmid), 133 fertile G0 adults yielded 17 (12.8%) independently transformed G1 progeny. The structure of FRT- and loxP-containing plasmids can be tested by recombination in bacteria, before one proceeds with Drosophila embryo injection (29,30). The arrangement of loxP and FRT sites in P[wFl] and P[SFl] suggests other potentially useful configurations. For example, one may reduce the size of a transgene-coplacement construct if upstream sequences are shared between the two genes to be compared, as when two cDNA sequences are driven by the same promoter. In such a case, the promoter may be placed upstream of the first loxP site and the two cDNAs placed in cloning sites 1 and 2. Because the loxP site is only 34-bp in length, it is minimally disruptive. Plasmids are now available (31) containing loxP and FRT sites in various combinations, thus facilitating the cloning of custom-made coplacement constructs. Quantitative analysis of expression differences between coplaced transgenes can make use of the paired-comparisons t test (or the equivalent two-way analysis of variance), which confers greater power than the standard t test (or one-way ANOVA). The power of the statistical test increases as the coefficient of variation of measurements across genomic positions decreases, and increases as the correlation in expression between transgenes at the same position in the genome increases. Qualitative analysis of transgene expression should precede quantitative analysis, because some genomic insertion sites may cause misexpression of the transgenes, and should therefore be excluded from the quantitative analysis. The “>” notation for FRT sites was introduced by Golic and Lindquist (3), so that the orientation of repeated FRT sites can readily be designated; “>X>” means that X is flanked by directly-repeated FRT sites, whereas “>X